syntactic reap.pt - inesc-id

84
Syntactic REAP.PT Cristiano Jos´ e Lopes Marques Dissertation for obtaining the Master’s Degree in Information Systems and Computer Engineering Jury President: Professor Jo˜ ao Ant´ onio Madeiras Pereira Advisor: Professor Nuno Jo˜ ao Neves Mamede Professor Jorge Manuel Evangelista Baptista Evaluation Jury: Professor Bruno Emanuel da Gra¸ca Martins 2011

Upload: others

Post on 11-Feb-2022

5 views

Category:

Documents


0 download

TRANSCRIPT

Syntactic REAP.PT

Cristiano Jose Lopes Marques

Dissertation for obtaining the Master’s Degree inInformation Systems and Computer Engineering

Jury

President: Professor Joao Antonio Madeiras PereiraAdvisor: Professor Nuno Joao Neves Mamede

Professor Jorge Manuel Evangelista BaptistaEvaluation Jury: Professor Bruno Emanuel da Graca Martins

2011

Acknowledgements

First of all I would like to thank Prof. Nuno Mamede and Prof. Jorge Baptista for providing

me with this opportunity and for his excellent guidance, support, motivation, and knowledge

sharing. I would also like to thank Andre Silva and Joana Barracosa, for their optimistic support

and for the many conversations, work meetings and helpful insights along the course of this work.

Their help and collaboration strongly improved the journey. Furthermore, my thanks go to every

author for their work and contribution with inestimable background knowledge. To my friends,

a special thank you, for helping and providing their own knowledge and expertise, even when it

was not related with this work. To Fernando Pardelhas, Pedro Ruivo and Joao Dias, for their

continuous help, valuable feedback, support, motivation and and dedication firmly encouraged

me to move on. Lastly, but definitely not least, to my parents and family, a special thank you

very much for all these years of comfort and encouragement that made this work possible.

This work was supported by project CMU- PT/HuMach/0053/2008, sponsored by FCT.

Lisboa, 2011

Cristiano Jose Lopes Marques

To my parents, Jose and Cidalina.

To my brother, Nuno,

Resumo

Ensino da lıngua assistido por computador e uma area de pesquisa que se foca no desen-

volvimento de ferramentas que melhorem o processo de aprendizagem da lıngua. REAP.PT

(REAder-specific Practice Portuguese) e um exemplo destes sistemas, em que o objectivo e aju-

dar a ensinar a lıngua portuguesa de uma forma apelativa, abordando temas que sao do interesse

dos utilizadores.

Esta tese foca-se na geracao automatica de exercıcios sintacticos e de vocabulario. Para

tentar cumprir esses objectivo, um conjunto de tarefas foram desenvolvidas: um estudo de

formatos de questoes pertinentes e quais os sistemas que criavam estes exercıcios; um estudo

de sistemas CALL com exercıcios sintacticos e semanticos; e modelacao/desenvolvimento de

um solucao arquitectural para o problema, incluindo diferentes recursos: filtros, classificadores,

tabelas com informacao lexical adicional e geradores automaticos de distractores.

Outro objectivo deste trabalho e incluir este modulo no sistema CALL - REAP.PT.

Abstract

Computer-Aided Language Learning (CALL) is an area of research that focuses on devel-

oping tools to improve the process of learning a language. REAP.PT (REAder-specific Practice

Portuguese) is a system that aims to teach Portuguese in an appealing way, addressing issues

that the user is interested in.

This thesis focuses on the topic of automatic generation of syntactic and vocabulary ques-

tions. In trying to successfully accomplish this goal, there are several tasks that were developed:

a study on the format of pertinent questions and which existing systems create these exercises;

a study of some CALL systems with syntactic and semantic exercises; and the modeling of an

architectural solution of the problem, including other resources such as filters and automatic

distractors generation.

Another goal of the present work is to be able to include this feature in the tutoring system

– REAP.PT.

Palavras Chave

Keywords

Palavras Chave

Ensino da lıngua Assistido por Computador

Geracao Automatica de Exercıcios

Exercıcios Sintacticos

Exercıcios Semanticos

Geracao Automatica de Distractores

Lıngua Portuguesa

Keywords

Computer Assisted Language Learning

Automatic Exercise Generation

Syntactic Exercises

Semantic Exercises

Automatic Distractors Generation

Portuguese

Contents

1 Introduction 1

1.1 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Requirements of Syntactic REAP.PT . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.3 Computer-Aided Language Learning . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.4 Structure of this Document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 REAP 7

2.1 Architecture of REAP.PT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 State of the art 13

3.1 Syntactic-Semantic Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.1.1 Crosswords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.1.2 Correspondence Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.1.3 “True” or “False” exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.1.4 Multiple choice exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.1.5 Fill in the blank exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3.1.6 Word Sorting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1.7 Short Answer Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.1.8 Open Answer Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.1.9 Alphabet Soup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.2 Automatic Exercise Generation Systems . . . . . . . . . . . . . . . . . . . . . . . 21

i

3.2.1 Working With Real English Text . . . . . . . . . . . . . . . . . . . . . . . 22

3.2.2 TAGARELA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.2.3 ALPHEIOS PROJECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4 Our Approach 25

4.1 Mahjong Lexical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.1.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

4.2 Choice of mood in subordinate clause . . . . . . . . . . . . . . . . . . . . . . . . 32

4.2.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

4.3 Collective Names and Nominal Determinants . . . . . . . . . . . . . . . . . . . . 34

4.3.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4.4 Teacher Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4.4.1 Mahjong Lexical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4.4.2 Choice of mood in subordinate clause . . . . . . . . . . . . . . . . . . . . 38

4.4.3 Collective Names and Names determinative . . . . . . . . . . . . . . . . . 39

5 Evaluation 41

6 Conclusion and Future Work 49

Bibliography 55

A 5th Grade P–AWL 57

B Number of exercises for each nominal determinant 59

B.1 Collective names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

B.2 Nominal Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

C Initial Form 61

ii

D Questionnaire 63

iii

iv

List of Figures

2.1 Architecture of REAP.PT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.2 Vocabulary Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.3 Example of Cloze Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3.1 Crossword (taken from http://www.prof2000.pt/users/amsniza/) . . . . . . . . . 15

3.2 Correspondence (taken from http://www.prof2000.pt/users/amsniza/) . . . . . . 16

3.3 Fill in the blank exercises (taken from http://guida.querido.net/jogos/) . . . . . 18

3.4 Word Sorting (taken from http://guida.querido.net/jogos/) . . . . . . . . . . . . 19

3.5 Alphabet Soup (taken from http://cvc.instituto-camoes.pt/sopa-de-letras.html) . 21

4.1 Mahjong Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.2 HTML structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

4.3 Filters and Classifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

4.4 Mahjong Table and Mahjong Evaluation . . . . . . . . . . . . . . . . . . . . . . . 28

4.5 Mechanism to award points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

4.6 Filters’s Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

4.7 Example “ate” (until) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

4.8 Chunk Subclause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

4.9 Example of Exercise of Nominal Determinant . . . . . . . . . . . . . . . . . . . . 34

4.10 QUANTD dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4.11 Teacher Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

v

5.1 The system was quick in giving an answer . . . . . . . . . . . . . . . . . . . . . . 43

5.2 The system was easy to use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

5.3 The help menu contains useful information presented in a concise and clear manager 44

5.4 Which exercise did you like best . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

5.5 Error rate by exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.6 Which exercise was the easiest . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.7 System’s global appreciation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

C.1 Initial Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

D.1 Questionnaire - 1st Part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

D.2 Questionnaire - 2nd Part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

vi

List of Tables

1.1 Functional Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2 Non-Functional Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

3.1 Exercises and language skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

4.1 Collective names and nominal determinants . . . . . . . . . . . . . . . . . . . . . 36

5.1 Test Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.2 Mahjong Lexical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.3 Average test duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

A.1 5th Grade P–AWL (1st Part) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

A.2 5th Grade P–AWL (2nd Part) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

B.1 Collective names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

B.2 Nominal determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

vii

viii

1IntroductionThe field of Education is certainly one of the most important and discussed topics in the

world. During the past two decades, the exercise of spoken language skills has received increasing

attention among educators. According to the Lisbon Summit (Lisbon European Council 2010),

learning a foreign language is extremely important for the curriculum enrichment of any person

and is considered, by the same entity, as one of five key skills of a successful professional. With

the advancement of technologies for information systems, the use of CALL has emerged as a

tempting alternative to traditional modes of supplementing or replacing direct student-teacher

interaction, such as the language laboratory or audio-tape-based self-study. This context is the

starting point of the REAP1 project and also of the REAP.PT2 project.

1.1 Goals

The main goal of the REAP.PT system is to present to the students rich and authentic

study material (texts, exercises, etc) that is deemed interesting by the students (based on their

topics of interest), and adequate to their learning needs and current skills, thus being able to

advance their learning process.

The aim of this work is to develop a module of REAP.PT to help learning the syntactic-

semantic component of the language, through the development of automatic exercises drawn

from actual texts taken from the Internet, and according to the topic preferences of the student.

This module can use features already present in the REAP.PT system (see Section 2.1)

While the main focus of the original REAP project has been the acquisition of vocabulary,

the exercises to be present by this module of REAP.PT will focus too on aspects of syntax that

are especially problematic for students of Portuguese as a second language, for example, the

1http://reap.cs.cmu.edu (last visited in December 2010)2http://call.l2f.inesc-id.pt/ (last visited in December 2010)

2 CHAPTER 1. INTRODUCTION

placement of clitic pronouns, adjectives or adverbs, formation of interrogatives, the alternation

between direct speech and indirect speech, between active/passive sentence structure or nomi-

nalizations. Because the exercises are generated automatically, careful design of the generation

process is necessary in order to assure the linguistic adequacy and the relevance of the question.

Above all, the exercises automatically generated by the system should not present ambiguous

solutions. For example: in the case of gap-filling exercises, it is important that there is not more

than one possible solution so that the system does not invalidate the answer given by the user

just because this was simply not the expected input.

Additionally, it is necessary to create an interface for teachers and another for students.

The teacher will have permissions to see the answers given by students, save all the answers and

editing/viewing the exercises. The student interface will give him access to available exercises

and allow him to solve them. From a technical point of view, special care must be taken in

the development of both interfaces (teacher and student) so that it is not necessary to install

plugins on the client browser.

1.2 Requirements of Syntactic REAP.PT

This section defines the requirements of the problem, for adequate project planning and

management in the presence of changing requirements. Defining requirements precisely is im-

portant to identify what exactly should the system do. Requirements can be divided in two

groups (Rost 2005): functional requirements (see Table 1.1) and non-functional requirements

(see Table 1.2). The main requirements for this module of REAP are thus defined as:

1.3. COMPUTER-AIDED LANGUAGE LEARNING 3

Requirement Description

Automatic Exercise GenerationThe exercise generation must be done with-out human intervention

Different types of exercises

Clitic pronouns, adjectives or adverbs, for-mation of interrogatives, the alternationbetween direct speech and indirect speech,active/passive or nominalizations

Teacher permission Teacher can see students’ result

Students inputStudents can answer questions through abrowser

Table 1.1: Functional Requirements

Requirement Description

High AccuracyGenerated exercises don’t have ambiguoussolutions

SecurityDifferent permissions for teachers and stu-dents

Persistence Save students’ answers and results

UsabilityUsers do not need to have advanced knowl-edge in Technology

Table 1.2: Non-Functional Requirements

1.3 Computer-Aided Language Learning

Currently, CALL can be seen as an approach to language teaching and learning in which

computer technology is used as an aid to the presentation, reinforcement and assessment of

material to be learned, usually including a substantial interactive element. It also includes

the search and the investigation of applications in language teaching and learning (Hacken

2003). Gamper and Knapp (2002) define CALL as “a research field which explores the use of

computational methods and techniques as well as new media for language learning and teaching”

(Gamper & Knapp 2002).

However, various difficulties have to be overcome in order to developed CALL systems,

namely: the financial barrier resulting from the need of strong financial support required to

obtain human and technological resources; availability and access conditions to hardware and

software in the teaching institutions; and the reluctance – still quite strong – from many people,

both teachers and students, in using new technologies, specially in language learning.

4 CHAPTER 1. INTRODUCTION

One of the most important requirements for this CALL system is the ability to retrieve and

use texts with updated information, as well as to present correct and appropriate content.

This goal has already been achieved as a result form previous research (Marujo 2009) (Cor-

reia 2010).

The motivation for this requirements can be summarized as follows:

• Experiential learning: it is easier for the user to learn gradually from their mistakes;

• Motivation: generally speaking, the use of technology both inside and outside the class-

room, tends to make the learning process more interesting. However, certain design issues

may affect the way a particular tool creates influences motivation;

• Enhance student achievement: if texts shown to the student are to be selected according to

their topic preferences, it is then likely that their motivation be enhanced, thus improving

their learning performance;

• Greater interaction: computers can adapt to students; adapting to a student usually means

that the student controls the pace of the learning but it also means that the student can

make choices regarding the content and the manner of learning, skipping unnecessary

items or doing remedial work on difficult concepts; such control makes students feel more

competent in their learning (Okan );

• Completeness: a fully-fledged CALL systems provides the student the opportunity to

interact using one or more of the four basic languages skills (reading, writing, listening

and speaking).

1.4 Structure of this Document

The present thesis consists in 6 chapters, and it is structured as follows:

• Overview of the REAP.PT is described in Chapter 2

• Chapter 3 describes the work related to the theme of this thesis: Portuguese language

syntax exercises and automatic question generation systems;

1.4. STRUCTURE OF THIS DOCUMENT 5

• The solution proposed is presented in Chapter 4;

• Chapter 5 introduces how the system here developed will be evaluated;

• Chapter 6 presents the conclusion and future work.

6 CHAPTER 1. INTRODUCTION

2REAPREAding-Practice1 (REAP) project was initially developed by the Language Technologies

Institute (LTI) of Carnegie Mellon University2 (CMU). It is a system that implements the main

ideas of CALL and it is specifically focused on vocabulary learning.

The goal behind REAP, as described by Michael Heilman and Maxine Eskenazi “is to furnish

appropriate, authentic texts to students to help in reading and vocabulary learning. In REAP,

a student sees short reading passages that contain a number of words (usually ranging from two

to four) from his or her list of target words to be learned from context. The passages are Web

documents of about one to two pages in length, covering a wide variety of topics” (Heilman &

Eskenazi 2006). This reading is complemented by practice in series of automatically generated

exercises.

In April 2009, a Portuguese version of REAP - REAP.PT3 begun being implemented. The

target audience for this system is primarily students who want to learn Portuguese as a foreign

language.

2.1 Architecture of REAP.PT

In order to understand how the current work is integrated in the remaining modules already

developed within the REAP.PT project. This section describes the current system architecture

(see Figure 2.1) (Correia 2010)(Marujo 2009):

• Web Interface: Two interfaces, one for student and other for teacher have already been

implemented (Marujo 2009). These can be accessed through any web browser. The web

interface is also responsible for the communication between the database and the DIXI

1http://reap.cs.cmu.edu (last visited in June 2011)2http://www.cmu.edu/index.shtml (last visited in June 2011)3http://call.l2f.inesc-id.pt/ (last visited in June 2011)

8 CHAPTER 2. REAP

Figure 2.1: Architecture of REAP.PT

software (Paulo, Oliveira, Mendes, Figueira, Cassaca, Viana, & Moniz 2008). This software

is a system for automatic speech generation from text which uses the most advanced

techniques of speech production. It enables audio playback of text from any document

(text-to-speech). An example of a direct application of this software is the possibility of

listening to words that are marked as focus word (words considered important and that

students should learn) or, for that matter, of any words in the text, sequences of words,

phrases and even entire sentences;

• REAP.PT Database: It consists of two databases. The first one is used to maintain the

state of the system during execution and use (profile’s information of each user, keeping

track of exercises for each user, questions that will be solved, etc). The second database

contains references to the Andrew File System (AFS) (Howard, Kazar, Menees, Nichols,

Satyanarayanan, Sidebotham, & West 1988) which is responsible for storing the content

of the lexical resources (like list of words, dictionary, texts) (Correia 2010);

2.1. ARCHITECTURE OF REAP.PT 9

• Nutch Crawler: When building ClueWeb094 (see below), a crawler was necessary. In

this case, the Nutch Crawler was used, wich aims at indexing and automatically browse

the World Wide Web, and extracting updated information (in this case, texts) (Moreira,

Michael, da Silva, Shiloach, Dube, & Zhang 2007);

• ClueWeb09: Is a collection of over 1 billion web pages (5 TB compressed, 25 TB uncom-

pressed) created by the LTI at CMU. This corpus contains texts in 10 different languages

(like Arabic, English, Portuguese or Spanish), compiled for the research on speech and lan-

guage technology. In the specific case of the Portuguese section of this corpus, it includes

37,578,858 pages, retrieved in 2009. This subset of documents from ClueWeb09 (about

160 GB compressed) constitutes the corpus currently being used in the REAP.PT project;

• Filter Chain: In spite of its size, only a subset of this corpus of texts is selected for use

in REAP.PT, if they don’t violate the following constraints or filters: text cannot have

less then 300 words (short texts are excluded); HTML format; only texts with words from

the Portuguese Academic List (P–AWL, see below) (Baptista, Costa, Guerra, Zampieri,

Cabral, & Mamede ) are kept; texts with profanity words are rejected;

• Topic Classifier: The texts considered valid by the filtering system are classified according

to topic and reading level, and them stored in the database (Correia 2010);

• Readability Classifier: Texts are classified according to the Collins-Thompson and Callan

algorithm (Collins-Thompson & Callan 2005);

• Question Generation: This module is responsible for creating the exercises that are given

to students at the end of each reading – definition question, synonym/antonym question,

hyperonym/hyponym question, cloze question and open cloze question.

• P-AWL: Ensures the quality of the vocabulary that students have to learn. To achieve this

objective, the REAP.PT use Portuguese Academic Word List (P-AWL), which can be de-

fined as a “general purpose vocabulary, with current (but not colloquial) words, which has

been designed for immediate application on a CALL web-based environment, currently de-

voted to improve students’ reading practice and vocabulary acquisition” (Baptista, Costa,

Guerra, Zampieri, Cabral, & Mamede ).

4http://boston.lti.cs.cmu.edu/Data/clueweb09/ (last visited in December 2010)

10 CHAPTER 2. REAP

• Infopedia5, PAPEL6 (Oliveira, Santos, Gomes, & Seco 2008), TemaNet7 and MWN.PT8:

Are resources used to implement the automatic question generation system.

In the next section, some of these components of the REAP.PT are presented in greater

detail.

Portuguese Academic Word List

It is important that a system meant for language learning ensures the quality of the vo-

cabulary that students have to learn. To achieve this objective, the REAP.PT use Portuguese

Academic Word List (P-AWL), which can be defined as a “general purpose vocabulary, with cur-

rent (but not colloquial) words, which has been designed for immediate application on a CALL

web-based environment, currently devoted to improve students’ reading practice and vocabulary

acquisition” (Baptista, Costa, Guerra, Zampieri, Cabral, & Mamede ). P-AWL consists of 1823

lemmas (and over 33k inflected forms associated to them). This list defines the focus words that

students should learn.

To implement the automatic question generation system, the following resources have been

used: TemaNet9, PAPEL10 (Oliveira, Santos, Gomes, & Seco 2008) and MWN.PT11. Another

feature important is the dictionary of Porto Editora, the Infopedia12 that contains 920,000

entries.

Language Proficiency and Preferences

The level of language proficiency of the students is calculated through the resolution of a

set of exercises presented to students when they are first introduced to REAP.PT. Based on this

level, each student is assigned a set of P-AWL words that they must learn when interacting with

5http://www.infopedia.pt/ (last visited in December 2010)6http://www.linguateca.pt/PAPEL/ (last visited in December 2010)7http://www.instituto-camoes.pt/temanet/ (last visited in December 2010)8http://mwnpt.di.fc.ul.pt/index.html (last visited in December 2010)9http://www.instituto-camoes.pt/temanet/ (last visited in December 2010)

10http://www.linguateca.pt/PAPEL/ (last visited in December 2010)11http://mwnpt.di.fc.ul.pt/index.html (last visited in December 2010)12http://www.infopedia.pt/ (last visited in December 2010)

2.1. ARCHITECTURE OF REAP.PT 11

the system (Correia 2010). There is a direct correlation between the current level of student

and the keywords is supposed to learn.

When the students enter their personal area, they have the opportunity to choose which

topics they find more or less interesting, so that later the texts and exercises presented by the

system may be more suitable to their topic preferences.

Listening Module

Listening comprehension is a complex but critical process to the development of the skills

necessary to learn a foreign language. This module in particular can be viewed as the result

of integrating two components: Software Text-to-Speech and Multimedia Documents. The first

component is the software DIXI (see Section 2.1), through which students have the opportunity

to hear the words or word sequences (phrases) that they have previously selected. According

(O’Malley, Chamot, & Kupper 1989) listening helps to discriminate sounds, understand the

vocabulary, acquire grammatical structures, and to interpret intentions. All these factors are

responsible for the preeminence of listening comprehension, especially in the early stages of the

language acquisition (Stern ).

The second module, Multimedia Documents, complements the speech synthesis software. A

type of multimedia documents presented in REAP.PT and that has been already implemented

is the digital talking books. These are an excellent tool to improve language proficiency since

it allows students to follow the oral reading of a text by a nature speaker, thus being able to

associate the sounds to words, learn prosody and other speech features.

Another feature that is available in Listening Module is the broadcast news (BN) stories,

which consists of news subtitles from television through an automatic speech recognition system

(ASR). ASR, the AUDIMUS (Meinedo, Caseiro, Neto, & Trancoso 2003) “combines the temporal

modelling capabilities of Hidden Markov Models with the pattern discriminative classification

capabilities of multilayer perceptrons”. In short, AUDIMUS converts an acoustic signal to text,

allowing the processing of sound produced by one or more speakers (Meinedo & Neto 2003).

The news stories are presented in small sized videos (one of news piece) along with the subtitles

produced by AUDIMUS.

Although it is not the main objective of this module, the use of BN can also help students

12 CHAPTER 2. REAP

learn facial expressions, gestures and emotions, witch are essential elements of oral, face-to-face

communication. This is a side effect of the visual training implicit in the BN, which becomes

especially important for people with hearing difficulties, but who have the ability to interpret

facial expressions and lip movements.

Exercises

The development of (semi-)automatically built exercises has been the focus of much research

(L2F) in the improvement of the REAP.PT system. Initially a set of cloze questions have been

manually built by a team of teachers and linguists (Marujo 2009). Later on, automatic generation

of questions has been introduced (Correia 2010) like vocabulary exercises (see Figure 2.2) and

cloze question (see Figure 2.3) cases.

Qual das seguintes e uma definicao da palavra criar?; Which of the following is adefinition of the word create?

a) conceber algo original; design something originalb) coser a orla de um tecido para que nao se desfie; sew the hem ofa cloth so as not to shredc) temperar em vinha-d’alho com o intuito de aromatizar a comida,conserva-la e torna-la mais terra; had garlic wine in order to flavorthe food, keep it and make it more landd) causar a morte; cause death

Figure 2.2: Vocabulary Exercise

O livro e tao extenso e tao complexo que o editor decidiu que nao era para umtradutor e contratou dois profissionais para o fazer.; The book is so extensive andso complex that the editor decided it was not for a translator and hired twoprofessionals to do it.

a) palestra; lectureb) assunto; subjectc) receita; reciped) tarefa; task

Figure 2.3: Example of Cloze Question

The goal of this thesis is to continue the work already done, including more types of exercises

in REAP.PT.

3State of the art

3.1 Syntactic-Semantic Exercises

The literature review in this area is extremely important to ensure that the implemented

exercises are consistent with what is taught and practised by Portuguese students in their first

years of school. Thus, it will be ensured that this thesis focuses on the really important and

interesting exercises for learning the language.

In Table 3.1, a match is made between the skills explored and the exercises’ formats that

are presented in this thesis. These exercises are available in format in digital format (such as

software applications, online exercises) and on paper (such as grammars, books).

Skills were divided into four groups:

• Reading Comprehension - Ability to understand and interpret what you read;

• Syntactic knowledge of language - Ability to apply the language rules to construct sen-

tences;

• Semantic Knowledge of Language (Lexicon) - Knowledge of word meaning;

• Meta-Linguistic Knowledge - Knowledge about the language used to describe the natural

language.

It is important to mention that with a small amount of imagination, each exercise format

can explore other skills that are not represented in Table 3.1. Essentially this table reflects the

skills exercised by the systems presented in this work.

In the following sections, several syntactic exercises are presented.

14 CHAPTER 3. STATE OF THE ART

Reading Com-prehension

Syntacticknowledge oflanguage

SemanticKnowledgeof Language(Lexicon)

Meta-LinguisticKnowledge

Crosswords X X X

Fill-in-the-blanks

X X X

Correspon-dence Exer-cise

X X

”Right” or”Wrong”

X X

MultipleChoice

X X X

Short Answer X X

Word Sorting X X X

Identificationof Grammati-cal Category

X

Open Re-sponse

X X X

AlphabetSoup

X

Table 3.1: Exercises and language skills

3.1.1 Crosswords

Crosswords (see figure 3.1) are usually done on paper, but there are also computer versions.

Words must be discovered (remembered or guessed) following tips provided by the exercise.

The most common use of crosswords consists in identifying a word from its definition (e.g.

“profissional da area de ensino”/“professor” ; professional in the field of education/teacher).

Other tips may involve metalinguistic knowledge (e.g “verbo casar no gerundio”/“casando” ;

verb married in the gerund/married).

There is already a system (Aherne & Vogel 2006) developed by the Computational Lin-

guistics Group – University of Dublin1 that automatically generates crossword puzzles, using a

pre-installed database with the target clues/words pair needed to these exercises. This database

was obtained with the support of the information contained in WordNet (Stark & Riesenfeld

1http://www.scss.tcd.ie/disciplines/intelligent systems/clg/clg web/ (last visited in December 2010)

3.1. SYNTACTIC-SEMANTIC EXERCISES 15

Figure 3.1: Crossword (taken from http://www.prof2000.pt/users/amsniza/)

1998). Besides these systems, there are tools that do not allow the automatic generation, but

nevertheless help the user in the creation of the exercise in a digital format, such as the Jcross2,

Instant Online Crossword Puzzle Maker3 or the Crossword Generator for Teachers4, where the

clues and the words have to be inserted manually into the system by the user. There are other

systems that search the web for definitions of words and then through a program of constraint

satisfaction (CSP) constructs the puzzle (Anbulagan & Botea 2008).

Some characteristics of this exercise format:

• difficulties on producing automatically the word grid;

• the ludic aspect oh the exercise is a motivator;

• students may take advantage from the crossing of words to guess the missing words so

indirect help way be available and direct assessment of language skill being tested may be

hindered.

3.1.2 Correspondence Exercises

Correspondence exercises consist of two columns of items that are to be aligned by appealing

to a single linguistic relation (meaning, syntactic transformation, etc.). They aim at simulating

the student’s ability to differentiate between several concepts being presented, or to systematize

2http://hotpot.uvic.ca/ (last visited in December 2010)3http://www.varietygames.com/CW/ (last visited in December 2010)4http://www.theteacherscorner.net/printable-worksheets/make-your-own/crossword/crossword-puzzle-

maker.php (last visited in December 2010)

16 CHAPTER 3. STATE OF THE ART

the concepts involved in the linguistic relation under study. One of the most common uses of this

exercise format is to make the correspondence between the word and its definition. (example:

phone/mobile phone).

For example, in Figure 3.2, a correspondence exercise is shown, aimed at drilling the pro-

nounning of different noun phrases, in several syntactic functions (subject, object, beneficiary,

etc.).

Figure 3.2: Correspondence (taken from http://www.prof2000.pt/users/amsniza/)

To create exercises in this format there is JMatch, although it does not manage the exercise

automatically, it only allows teachers to build these exercises in digital format. Regarding the

(semi-)automatic generation:

• working the other way around (from pronouns to full-hedged Prepositional Phrases or

Nominal Phrases) raises the problem of anaphora resolution (Carbonell & Brown 1988), a

natural language processing (NLP) task that still has a limited success.

• the exercise may be automatically generated, depending on the more or less complexity

on the relation being tested. In this case, automatic identification of Nominal Phrase

(NP)/Prepositional Phrase(PP) and their function is likely to produce incorrect results

because, among other factors, the yet unsolved problem of PP attachment and the conse-

quent incomplete delimitation of the constituent being pronominalized.

3.1. SYNTACTIC-SEMANTIC EXERCISES 17

3.1.3 “True” or “False” exercises

True-or-false exercises are often used in text comprehension. They have the great advantage

of being aplicable in many other areas of language learning as well. For example, agreement

between nouns and adjectives can be tested, given the following instruction/example pair:

“Marca “correcto” or “incorrecto” as frases seguintes: Sempre julguei competente o gerente

e a directora” (Mark as “correct” or “incorrect” the following sentences: I have always considered

competent the manager and the director).

The main disadvantage of this type of exercises is the loss of specificity for each question,

since it may not reveal exactly the knowledge that has been activated to answer.

The most common systems for creating this type of exercise have a similar architecture to

(Bastos, Berardi, & Silveira 2004) where there is a database with a set of questions and answers

previously defined by a teacher.

3.1.4 Multiple choice exercises

Multiple-choice or cloze questions is probably the most widely used type of automatically

generated exercises. In their most basic form, they consist of a sentence (the stem), which is

provided with a blank space from where a word or string of words has been removed; and a set of

possible answers. Usually, the correct word is randomly placed among several incorrect choices

(distractors or foils) and while ideally only one possible solution exists, the set of distractors

are carefully designed to engage different linguistic strategies in order to pinpoint the language

knowledge that the student is supposed to have acquired and is being tested for. Much research

has already been devoted to the theoretical, psychological, cognitive and educational aspects

involved in the making of cloze question (Armand 2001), while several systems have been built to

automatically generate cloze question (Correia 2010) (Aldabe, de Lacalle, Maritxalar, Martinez,

& Uria ).

The distractors have a very important role in the automatic generation of these exercises.

Distractors can be extracted from several sources. They can be, for instance, random words,

phonetically similar words, antonyms or variations of the correct word (Correia 2010).

There are already some solutions in REAP.PT and ArikIturri

18 CHAPTER 3. STATE OF THE ART

(Aldabe, de Lacalle, Maritxalar, Martinez, & Uria ). The first system, focuses pri-

marily on vocabulary and is based on distractors that were implemented using a

set of existing resources (like PAPEL, TemaNet and MWN.PT). The second system

“uses techniques of replacement and duplication of declension cases (inessive, abla-

tive, dative...), number (singular/plural) or the inflection paradigm (finite, indefinite)”

(Aldabe, de Lacalle, Maritxalar, Martinez, & Uria ).

3.1.5 Fill in the blank exercises

This type of exercise consists of blank spaces and words. The student has to match the

words with the blank spaces. The linguistic context of the blanks consist of real utterances so

they work both as clue and as conveyor of language knowledge.

Another version of this exercise is when students instead of having the words to fill the

blank spaces, only have tips for each blank space. Both versions are a powerful way to practice

various syntactic aspects of the language.

In Figure 3.3, an example of this exercise format can be seen, where each space has to be

filled by an adjective.

Figure 3.3: Fill in the blank exercises (taken from http://guida.querido.net/jogos/)

Another version of this exercise may be used for learning word inflection, for example, the

student only knows which is the verb to apply. The verbal agreement, mode and tense are

implicit in the exercise, e.g. “O Pedro queria que o Joao (ir) a Lisboa” / “Peter wanted

that John (go) to Lisbon”.

JCloze5 is a tool that facilitates the creation of these exercises in digital format, but it

requires that all information is entered by the user.

5http://hotpot.uvic.ca/ (last visited in December 2010)

3.1. SYNTACTIC-SEMANTIC EXERCISES 19

(Semi-)automatically generation of this type of exercises, is difficult since seldom is a word

so narrowly constraint in its use that its use that exactly same syntactic slot; secondly, automatic

slotting of syntactic structures is often too imprecise to be effective (except for simple cases)

(Correia 2010) and (Curto 2010).

This exercise format can shares several characteristics with the previous format: the gen-

eration of wrong answers for each space to be filled can also be generated by distractors; the

systems presented above may also be used to generate this exercise format.

There are other systems (Goto, Kojiri, Watanabe, Iwata, & Yamada 2010) that use a slightly

different approach from the previous ones:

• the system extracts sentences based on machine learning;

• estimates a blank part based on discriminative model;

• Generates distractors based on statistical patterns of existing question.

3.1.6 Word Sorting

These exercises consist of sentences whose words have been scrambled so that the student’s

may drag-and-drop them in the correct order.

Figure 3.4: Word Sorting (taken from http://guida.querido.net/jogos/)

The easiest way to generate these exercises automatically is to take a phrase from the corpus

and randomly change their word order. The difficulty is in assessing the response because the

sentence can be constructed in different ways and still be right.

Despite not having found any system that generates this format of exercises, this is exten-

sively used in text books and language materials to emphasize students’ grasp of the meaning

and the relationships between words.

20 CHAPTER 3. STATE OF THE ART

3.1.7 Short Answer Exercises

This type of questions is used to test the knowledge of different fields, making questions

where the answer is just a word or a small set of words. The great advantage of these exercises

is that instead of the exercises presented above, these are the ones where the answer depends

only on the students’ knowledge, because they do not contain any help. But one disadvantage,

is the great difficulty in carrying out the automatic correction because the student can write the

answer in different ways, including with spelling errors that do not necessarily indicate that the

student does not know the answer.

These exercises are not usually associated with syntax exercises, al-

though they are still used in some fields: singular/plural, masculine/feminine

(Costa & Traca 2010) (Costa & Mendonca 2010) (Seixas & Castanheira 2010). Exam-

ple: “Qual o plural de imagem?” (Costa & Traca 2010).

There are already some applications that allow teachers to create short-answer tests in digital

format (Jquiz), but in these cases teachers must first insert the questions and the corresponding

answers. (Semi-)Automatic generation systems available for this exercises’ format (Siddiqi,

Harrison, & Siddiqi 2010) are based on syntactic and semantic features, which introduced by the

processing chain, and in accordance with a set of preprogrammed rules, are generated questions.

3.1.8 Open Answer Exercises

Exercises where there are no restrictions on possible answers to a question. Each student

structures the response in the way that seems to be more appropriate. It may be exercises that

require some imagination (example: “Write a text that uses direct speech” ) or direct questions

about a specific topic.

For this reason, the automatic generation of these exercises is more complex. Nevertheless,

there is already some work done in this field (Sukkarieh, Pulman, & Raikes 2003) (Pulman &

Sukkarieh ), using the following tools:

• Modules to normalize students’ answers;

• Neural Networks and Machine Learnings, which attempt to calculate the similarity ratio

between the student’s answer and the correct answer.

3.2. AUTOMATIC EXERCISE GENERATION SYSTEMS 21

3.1.9 Alphabet Soup

This game can be used as an exercise. The goal is to the students find the answer to certain

questions or tips that are given along with to the alphabet soup. This exercise combines learning

with entertainment to teach the language.

Figure 3.5: Alphabet Soup (taken from http://cvc.instituto-camoes.pt/sopa-de-letras.html)

The difficulty of creating the table is one of the biggest drawbacks in the automatic gener-

ation of this exercise. Subsequently, there are other drawbacks:

• People with dyslexia can not do these exercises;

• The time to resolve the exercise is not related to the linguistic profienciency.

JClic6 is a tool that can create alphabet soups in digital format. The big drawback is that

all the information needed to create this format exercise has to be entered manually.

3.2 Automatic Exercise Generation Systems

Currently there are several tools which aim is to help students/teachers in their lessons.

Systems performing exercises generation are usually domain-specific because of the complex-

ity of the operations required for this task. For example, in the domain of Mathematics (Costa

2003) it is possible to mention the Computer Aided Generation and Solving of Math Exercises

6https://launchpad.net/jclic/ (last visited in December 2010)

22 CHAPTER 3. STATE OF THE ART

(Tomas & Leal 2003), AGILMAT (Tomas, Leal, & Domingues ) and others; in the area of music

exercises, there is the i-maestro(Ng 2008); and in the domain electric ac circuits analysis

(Cristea & Tuduce 2005), adpatative e-learning system developed by

(Holohan, Melia, McMullen, & Pahl 2006). System that include automatic generation of

syntactic exercises are not very common but the offer, if limited, already exists, especially for

English. In the next sections, some of these systems will be presented.

3.2.1 Working With Real English Text

The Working With Real English Text (WERTI) system (Dimitrov 2008), presents several

features that are interesting for the purpose of this work.

This system incorporates a chain of NLP, which filters real texts in order to generate various

syntactic exercises for English. In summary, this processing is divided into three phases:

• HTML processing (Pre-Processing): Removes the pages’ contents that surrounds the text

(tags, links), returning only the relevant text

• Linguistic Processing: responsible for tokenization, sentence boundary detection and part-

of-speech tagging.

• Enhancement Processing (Post-Processing): Performs the necessary operations to the cre-

ation of the exercise (it is different for each type of exercise).

The exercises produced by this system are generated with the help of a set of rules imple-

mented in the NLP chain, for example:

Gerunds vs Infinitives – In the presence of certain prepositions, it is required the application

of the gerund. These rules are in NLP grammar and they hide the tenses so that later students

can fill them in adequately - the verb to use is shown next to each space.

Wh-Questions – 126 standard rules are identified to formulate questions from the texts.

While these exercises way appear as too simple for a native speaker, they correspond to

syntactic topics where second language learners and known to have difficulty in learning English.

Color Different colors identify the words that match with the grammatical category that it is

being practised. These colors are accompanied with a caption.

3.2. AUTOMATIC EXERCISE GENERATION SYSTEMS 23

Click Students click on the words that they think can match to the grammatical category. If

it is correct the word turns green, otherwise it turns red.

Multiple Choice For each blank space in a text, several words of the same grammatical are

given, and the student has to choose the right one.

Category In the case of determinants, gerunds vs. infinitives, phrasel verbs and prepositions,

blank spaces appear so that students fill them properly. In the Wh-Question case, the

student must arrange the words given by the system to create a question.

3.2.2 TAGARELA

TAGARELA System (Bailey & Meurers 2008) is an intelligent computer-assisted language

learning that provides opportunities to practice reading, listening and writing skills. This sys-

tem can be accessed through a web browser and has the ability to offer individualized feedback

on speeeling, morphological, syntactic and semantic errors. Architecture of this system is di-

vided into two main components, which aim to make the syntactic and semantic analysis of the

students’ answers. The first component – Form Analysis – consists by four modules:

• Tokenizer: have special attention to cliticization, contractions, abbreviations;

• Lexical/Morphological Lookup: returns multiple analysis based on CURUPIRA lexicon

(Martins, Nunes, & Hasegawa );

• Disambiguator: set of rules for disambiguation;

• Parser: syntactic analyzer ascending.

The second component of the architecture – Content Analysis – tries to establish a semantic

agreement between the student’s answer and the correct answer.

This architecture supports different formats of exercises: fill gaps, vocabulary, description

of pictures and text description of the student hears.

24 CHAPTER 3. STATE OF THE ART

3.2.3 ALPHEIOS PROJECT

This software’s7 aim is to support the learning of two extinct languages - Latin and Ancient

Greek. To do so, this application provides: word definitions, word morphology, links to sections

of grammar considered relevant, manager of learned words and words left to be learned by each

user, automatic translator.

Regarding syntactic exercises, the application provides a specific tool - “My Diagrams”:

when a sentence is selected, this tool allows the student to create a dependency tree, analysing

each word according to its syntax; it also provides a set of questions for which the student must

classify each POS that is shown.

Although this software is being used as a browser plug-in, and the learning is being made

through the text of web pages, Alpheios also provides enriched texts, which allows the full use

of all the potential of this application.

Briefly, all systems presented in this work are composed of a processing chain that en-

riches the text with syntactic and semantic information. This information is useful for

generating exercises at a later stage. The major constraints consist of a limited number of

different types of exercises that are available on each system and also in small percentages of

exercises with errors due to problems with semi-automatic generation.

7http://alpheios.net/ (last visited in December 2010)

4Our Approach

To achieve the goals previously defined, it is expected that the developed module can gen-

erate three different types of exercises, wich are presented in this chapter.

4.1 Mahjong Lexical

In these exercises the student has to make a correspondence between the lemma and the

definition of a word.

This section describes the solution architecture adopted to automatically generate these

exercises (see Figure 4.1).

Figure 4.1: Mahjong Architecture

The words and their definitions will be taken from the Porto Editora dictionary. To represent

words and their definitions, the dictionary uses an HTML structure based on <span> and <div>

that group inline-elements and block-elements. Figure 4.2 shows an example of the HTML

structure for the word correr “run”, where each grouping <span> indicates a grammatical

category (transitive verb, intransitive verb) and also indicates the origin of the word (from

26 CHAPTER 4. OUR APPROACH

Latin). Each <span> element is composed ofs several <div> tags that group all information

needed (e.g. definitions).

Figure 4.2: HTML structure

The second step is to filter the definitions (see Figure 4.3), since only a subset of the

definitions is going to be classified later.

The filters’ module performs the following selection procedures:

• Cognate Words - Ideally, definitions containing words cognate to the target word should

be discarded, since this an obvious cue to the correspondence between a lemma and a

definition. As an approximation to this concept the Levenshtein distance was used, where

the algorithm implemented in this filter verifies if there is any word in the definition with

a high level of similarity with the target word. For example, for the target word “escrever”

(to write) the following definition is rejected “passar a escrito” (passing to the writing)

since “escrito” is cognate to “escrever”;

• Length - Only definitions of more than one word (to avoid similarities with synonyms’

exercises) and less than 150 words (to avoid very long definitions) are considered;

• Characters - characters that might hinder the understanding of a given definition are

removed. Examples: numbering of definitions, semicolon, cardinal, etc.

4.1. MAHJONG LEXICAL 27

Figure 4.3: Filters and Classifiers

A set of classifiers is also applied to the definitions:

• Difficulty - Usually, basic and more current meanings are shown first in dictionary’s defi-

nition, while more specialized or rarer meanings appear later.

For example: Target word – “caneta” (pen)

1. Beginners level - “utensılio que contem tinta e serve para escrever ou desenhar” (tool

that contains ink and is used to write or draw)

2. Intermediate level - “tubo ou cabo que se encaixa um aparo para escrever ou desenhar”

(pipe or cable that fits a nib to write or draw)

3. Advanced level - “especie de cabo ou pinca com que segura um cauterio” (kind of

cable or clamp that holds a cautery)

• Specific Domain - several definitions are very specific to a given field or profession. These

are classified as “advanced level” since they requires a deeper knowledge of the language.

Example: Target word – “relogio” (watch)

definition – BOTANICA termo com que se designam diversas malvaceas (BOTANICS

term with which several Malvaceae are designated)

The pre-processing steps here described are required to be performed once for each word,

and the results are stored in the REAP.PT database, in Mahjong table (see figure 4.4).

28 CHAPTER 4. OUR APPROACH

Figure 4.4: Mahjong Table and Mahjong Evaluation

The REAP.PT system is student-oriented, and the word-definition pairs are chosen by the

system developed in this work according to the student profile, therefore:

• Student’s level - influences the number of word-definition pairs that are presented to the

student; it also determines the difficulty level of the definitions presented;

• Student’s history - influences which words are selected by Syntactic REAP.PT to be present

to the student, showing words that the student doesn’t know yet. Words are considered

unknow if they were never searched by the student in the dictionary; or if the student

never matched them correctly with their definition; or if the students indicated that they

did not know them.

In each exercise of Mahjong Lexical, at least one pair word-definition will be repeated;

meaning that in each exercise there will be a word-definition pair that has already been presented

to the student in previous exercises. This feature serves to:

1. Avoid “forgetting” the definition of a new learned word, for it will be submitted periodically

in exercises;

2. Outwit cases where students hit a pair by luck and not through their knowledge;

3. Try to teach a new word, when students systematically miss the definition of that word.

Students’ performance before the repeated word-definition pair, will directly influence the rate

of students’ retention. The more times the student hit repeated words, the higher will be the

retention rate, otherwise it decreases. Only students with high retention rates in the words

repeated can claim to know the words’ meaning, while for students who only hit once, nothing

can be said about this learning aspect.

4.1. MAHJONG LEXICAL 29

Additionally, a scoring mechanism was added (see Figure 4.5). Initially a set of points is

given to the student, where each word-definition pair is worth 20 points. During the resolution

of the exercise, the student may be penalized according to following the set of rules: for each

wrong word-definition correspondence s/he loses 10 points; and for each minute passed without

the solution s/he loses 10 points with a maximum of 30 points.

Figure 4.5: Mechanism to award points

There will be also a word-definition pair with a difficulty level higher than the current

student’s level. If a student hit in this pair, they will receive 40 bonus points to their score.

The implementation of this scoring’ system tries to accomplish several objectives: to provide

greater feedback about student performance; to keep the student motivated in the exercise;

to prevent the student from solving the exercise by repeating his/her tries; and to provide a

“summary” of the student’s performance to the teacher.

The evaluation module, apart from the scoring’ system, also checks if the student hits/misses

a given word-definition correspondence and if the student has finished the exercise. The result

of this evaluation is stored in the REAP.PT database (see Figure 4.4) and it will be used later

to analyze the student’s learning progress.

The automatic correction of the exercise locks only the right answers, keeping the wrong

30 CHAPTER 4. OUR APPROACH

answers still selectable for further tries, and giving the student more opportunities to solve the

exercise correctly. Thus, we try to motivate the student to solve the exercise and to make

him/her learn the definitions he/she still doesn’t know.

4.1.1 Data

Figure 4.6: Filters’s Results

The results for this exercise are directly dependent on the words that teachers want to teach.

4.1. MAHJONG LEXICAL 31

For illustrative purposes, in this section we used the list of P-AWL words (5th grade) (see annex

A), that students have to know.

For the set of 167 words, there are 554 definitions available in Porto Editora dictionary (see

picture 4.6). From these definitions, only a subset of 439 definitions respect the “filter size”

and, of these, only 431 are considered by the filter that removes the definitions with invalid

characters.

Finally the definitions go through a third and final filter that removes definitions that contain

cognate words of the target word. In this last filter, 419 definitions are properly classified

according to their difficulty level. Completing this process, we get 222 definitions of basic level,

148 definitions of intermediate level, 49 definitions of advanced level and 135 definitions were

rejected from the initial set of 554 definitions (see Figure 4.6).

32 CHAPTER 4. OUR APPROACH

4.2 Choice of mood in subordinate clause

Learning the vocabulary of subordinating conjunctions and conjunctive phrases implies the

acquisition of syntactic restrictions imposed to the mode of the subordinate clause they intro-

duce. These constraints are also related with the tense and mood of the main clause. These

cloze questions allow students to practice reading comprehension while enhancing the syntactic

knowledge of the language. Thus, for example, “ate que” (until) imposes the infinitive or the

subjunctive mode, but does not accept the indicative mode (see Figure 4.7).

Figure 4.7: Example “ate” (until)

For the automatic generation of the exercise, the CETEMPublico Corpus (Santos & Rocha

2001) is used, after being processed by the STRING processing chain of L2F (Hagege, Baptista,

& Mamede 2010). A large set of conjunctions and conjunctive locutions have been registered in

the system lexicon. A very general set of rules creates a SC (subclause) chunk that links these

expressions to the first verb of the subordinate clause (see Figure 4.8). To run the filters that

generate the potential stems for this exercise, the distributed computing platform Hadoop is

used to process this large corpus.

Figure 4.8: Chunk Subclause

To generate the wrong answers, the L2F verbs’ generator (Hagege, Baptista, & Mamede

2010) was used. Furthermore, in order to avoid ambiguity in the wrong responses generated, we

added the following set of restrictions:

• if the verb is in the indicative (present, imperfect, perfect, pluperfect, future) — generate

verbs in subjunctive (present or imperfect);

4.2. CHOICE OF MOOD IN SUBORDINATE CLAUSE 33

• if the verb is in the subjunctive (present or imperfect) — generate verbs in indicative

(present, imperfect, perfect, pluperfect, future);

• if the verb is in the infinitive — randomly pick one of the tenses of the indicative or the

subjunctive;

• in the selection of distractors the homographs of the target word should not be used.

4.2.1 Data

To generate the stems for this exercise, the corpus CETEMPublico (Santos & Rocha 2001)

was used, which contains over 190 million words and 7 million setences. In order to extract

the candidate sentences in a fast and expedient way1, the Hadoop2 plataform for distributed

computing was used. This platform is based on MapReduce and uses a cluster to process

large amounts of data. In this way, was possible to extract 720,784 sentences and generate the

corresponding exercises of this type.

1Initial experiments showed a 95,6% reduction of processing time using Hadoop against a sequential executionof the processes. Even so, it took about 1h30m to process the entire corpus.

2http://hadoop.apache.org/ (visited in Jul. 2011)

34 CHAPTER 4. OUR APPROACH

4.3 Collective Names and Nominal Determinants

The purpose of this exercise is to learn the subtle distributional constraints observed between

a determinative noun (Dnom) and the noun it determines (see Figure 4.9). This exercise also

serves to teach the classifying relationship between collective names and common names, since

collectives often function as Dnom on common nouns.

Figure 4.9: Example of Exercise of Nominal Determinant

These exercises are generated from real sentences, taken from the corpus. In these sentences,

a quantifying dependency (QUANTD) (see Figure 4.10) has been extracted by the syntactic

parser XIP (Hagege, Baptista, & Mamede 2010). This dependency holds between a Dnom

functioning as the head of a nominal or prepositional phrase (NP or PP) and the head oh the

immediately subsequent PP introduced by preposition “de” (of).

Again, the distributed computing platform Hadoop is used to retrieve from the large-sized

corpus the sentences that can be potential stems for this exercise.

Figure 4.10: QUANTD dependency

A new set of lexical information was added to the lexicon in order to generate adequate

distractors for the exercises. To do that, a list of determinative and collective names was added.

A set of semantic features was defined and accorded to these words. These semantic features

consist on the following categories: Human, Animal, Food, Organization/Institutions, Object,

Nature, Military and Local. By using these features, potential distractors that share the same

traits as the target word are never select, thus avoiding to generate unwanted correct solutions

4.3. COLLECTIVE NAMES AND NOMINAL DETERMINANTS 35

as foils. An additional set of constraints was also added to prevent the generation of distractors

that, in a figurative use, would be possible solutions. For example, Dnom associated to the

“Animal” feature (alcateia, “pack”, speaking of wolves) are not used for human nouns (e.g.

uma alcateia de polıticos “a pack of politicians”) since this combination might be used in a

ironic formulation.

To make the exercise more interesting, in the generation of distractors the same number

and gender of the target word is kept.

One of the important points in the CALL system is that the application is able to provide

feedback to the student. When the student incorrectly chooses a given word-definition pair, s/he

is given some feedback by the system. This feedback consists of three parts:

• the definition of the wrong answer;

• a real sentence as an example of the correct use of the selected word;

• pictures of the word selected.

This feedback system gives the indication that the student missed a question, and tries to

teach the meaning of the given nominal determinant and also its proper use (giving examples of

images, definitions, and uses the correct given determinative noun).

The illustrative images of the nominal determinant that appear on the feedback system

were selected in advance and stored in the REAP.PT database. To select these images was

necessary to retrieve the first ten images available in google images for each determinative noun.

Subsequently, through a voting system, the two most voted images (in 10 possible) are selected

and used by the REAP.PT system.

Following the same logic of the Mahjong Lexical exercise, it is important that there is some

repetition in the nominal determinants presented to the student. The Syntactic REAP.PT

presents sentences that are different but whose correct nominal determinant used to fill the

sentence has appeared in previous exercises. If the student hits the nominal determinant in

different contexts it means that s/he knows its meaning and knows how to make a proper use

of it.

36 CHAPTER 4. OUR APPROACH

4.3.1 Data

The syntactic REAP was able to extract from the corpus approximately 19,000 sentences,

of which 46,61% involve determinative nouns exercises and 53.39% correspond to the collective

names (see table 4.1). It should also be noted that determinatives names that appear more

frequently (see annex B) in the exercises are “percentagem”, “pacote” e “pedaco” and the

most common collective name (see annex B) in the generated exercises are “equipa”, “album”

e “comissao”. Although some nominal determinants /collective names do not appear in the

correct answers generated by this module, they are used as a distractors.

Exercise Number of exercises %

Collective names 9,035 46.61%

Nominal determinants 1,0345 53.39%

Table 4.1: Collective names and nominal determinants

The Hadoop3 plataform was used to process the corpus with the filters in order to generate

the exercise, in a similar way as with the previous game.

3http://hadoop.apache.org/ (visited in Jul. 2011)

4.4. TEACHER INTERFACE 37

4.4 Teacher Interface

This section presents the work developed in the teacher interface for the REAP.PT system.

At this stage, we felt the need to create an interface that would be simple and intuitive but

at the same time could give all the information needed by the teacher in order to assess the

learning process of each individual student and the class in general. This interface assumes

particular relevance because if teachers feel they can not control the progress of the students

(given answers, questions presented, results, etc.) they will not use the system. To avoid this

situation it is interesting that the system, in addition to providing a general feedback about the

students’ performance, may also give all relevant data to the teacher if s/he wishes to validate

any particular information or make an assessment using her/his own criteria.

The pages that allow the visualisation of the students’ results are composed of two parts:

the first entitled “General Information”, where the teacher a general overview of the current

status of the student. The second, called “Detailed Information”, which contains all data of

all exercises solved by student, separated by exercise. Thus, teachers can see the evolution of

students over the use of Syntactic REAP.

4.4.1 Mahjong Lexical

In the General Information screen, teachers can view the summary of a student’s perfor-

mance in percentage (0% to 100%). This performance is the weighted value between the number

of points earned versus the maximum number of points that the student could obtain.

In order to complement and help understanding this performance, the percentage is pre-

sented in different colors: red (poor performance), orange (can improve), blue (successful out-

come) and green (good result). On this page the teacher can also check:

• number of exercises solved;

• time spent solving exercises;

• number of errors;

• words whose meaning the student already knows;

• words whose meaning the student does not seem to know.

38 CHAPTER 4. OUR APPROACH

In the Detailed Information screen the teachers can view the student’s history. For each

exercise, the teacher can check the performance of the student, the mistakes, the points obtained,

and the retention rate associated with each exercise. In order to be simpler to understand, these

results are presented in a graphic and a table format. The graphics (see figure 4.11) show

the oscillations of student performance and the retention rate. As in the General Information

screen, graphics are colored with: red (bad performance), orange (can improve), blue (successful

outcome) and green (good result).

Figure 4.11: Teacher Interface

4.4.2 Choice of mood in subordinate clause

In the Detailed Information screen the teacher can analyze the results of the student for

each exercise, namely:

• the number of errors committed;

• the answers submitted by the student;

• check if the student failed to respond properly to the exercise.

4.4. TEACHER INTERFACE 39

4.4.3 Collective Names and Names determinative

On the page General Information screen the teacher sees an overview of the student

performance. Here it is possible to view:

• number of exercises presented;

• number of errors;

• number of correct answers;

• determinative names / collective names that the student already knows how to use cor-

rectly;

• determinative names / collective names that the student still does not know how to use

correctly;

On the Detailed Information page the teacher can monitor the answers given by the

students in each exercise (the system indicates the correct answer for each exercise). It is also

possible to view the question presented in each exercise, the number of wrong answers, the

retention rate and the associated information if the student successfully solved the exercise.

40 CHAPTER 4. OUR APPROACH

5EvaluationTo assess the performance of the Syntactic REAP.PT, two groups of students were given

the system to try it and comment on its use. This evaluation consisted in a Syntactic REAP.PT

Session, composed of three parts:

• initial form (see annex C): tracing the students’ profile (e.g. age, native language, etc.);

• exercises: three mahjong lexical exercises are presented (one for each difficulty level), three

mood selection for subclause verbs exercises, three nominal determinants exercises, and

three collective names exercises;

• final questionnaire (see annex D): aimed at qualitative assessment of the system (e.g.

system case of use, which exercises are most relevant or appealing, etc).

The exercises generated by the Syntactic REAP require some knowledge of Portuguese as

they call upon more advanced language contents. In this way, to evaluate the system, it is

required that the subjects of the evaluation comply with this condition: if they are non-native

Portuguese speakers, they already should have an elementary knowledge of the language. Due to

the impossibility of gathering a group with the characteristics described in time for this exercise,

the tests were conducted on a group with relatively similar characteristics - Portuguese native

speakers in the 3rd and 4th grade. In order be able to contrast this group performance, the same

test was also performed with another group of native speakers with at least a college degree.

Test Users

Therefore, 45 subjects performed this exercise, 18 in Group 1 and 31 in Group 2. A brief

overview of these two groups in presented in Table 5.1. Naturally, while de age of Group 1

subjects is quite homogeneous, Group 2 varied from 19 to 70, the average age being 24. Most

subjects from Group 2 know other languages (such as english, french, spanish, italian and

42 CHAPTER 5. EVALUATION

german). The choice of Group 1 can be justified by their knowledge of Portuguese as mother

language but their still limited vocabulary - this being one of REAP.PT main learning targets.

GroupNumber ofusers

% Average age Other languages

1 18 37% 8 –

2 31 63% 24 english, french, spanish, ital-ian, german

Table 5.1: Test Users

As we can see in the table, the nationality of the group differs between the portuguese and

brazilian, and most of the people that belong to the Group 2 (with college degree) speak other

languages such as english, french, spanish, italian and/or german.

Assessment – Test Environment

The testing environment was different for each group. Group 1 did the exercises one at time

in the room’s computer, and a team member of REAP.PT project helped them to access the

starting page and occasionally had to explain the exercises or some unknown word since some

children still hadn’t fully mastered their reading skills. Each child took about 26 minutes to

complete the exercises and the questionnaire.

Users from Group 2 did the test at their own homes, without the aid of any element of the

REAP.PT project.

43

Figure 5.1: The system was quick in giving an answer

Results and Discussion

As shown in Figure 5.2, move than 77% of the users found the system easy to use. In this

way we can say that one of the goals of this project was achieved, namely, to create a system

with a simple and easy to use interface, so that all users (even some inexperienced users) were

able to use the Syntactic REAP.PT. In Table 5.3 we can see that only 39% needed to use the

“Help” button.

Figure 5.2: The system was easy to use

The pre-processing of the exercises is made in advance and stored in the REAP.PT database.

Thus, the generation stage, which is the most time-consuming step of the processing, is already

44 CHAPTER 5. EVALUATION

Figure 5.3: The help menu contains useful information presented in a concise and clear manager

done when the user access the system. This is why the system is so quick when responding to

requests from users as is a reflect in their answer to the question shown in Figure 5.1.

Figure 5.4: Which exercise did you like best

Regarding the results obtained in the three Mahjong Lexical exercises that were presented,

Group 2 obtained better results with a performance of 84% (standard deviation = 6,6%), while

the users from Group 1 made more errors and obtained a performance of 54% (standard deviation

= 12,4%).

Users from both group reported that the exercise they liked most (44%) was the Mahjong

Lexical (see Figure 5.4), the next most enjoyed exercise was the “Collective Nouns” (27%) fol-

lowed by “Nominal Determinants” (17%) and the “mood of subordinate clause” (12%). However

45

it is remarkable that the 2 Groups had clearly different impressions about these three exercise.

While Group 1 liked “Collective Nouns” as 2nd best ( 40%), Group 2 shows similar preference

for each of these three (16%, 19% and 21%) this way have to do with the that collective names

are an explicit grammatical subject at “escola primaria” (elementary school) thus contributing

do a higher familiarity with the topic which entailed this preference trend.

Another interesting aspect to note is the error rate for each exercise. Starting at Mahjong

Lexical, in Table 5.2, we can see that in both groups the more the difficulty level of exercises

increases, more mistakes the students do, which seems to confirm the strategy followed to dis-

tinguish these levels from the sequence of definitions inside the dictionary entry of each target

word (see Section 4.1)

Group 1 Group 2

Easy levelaverage=1.07 errors/exercise,standard deviation=0.98,mode=1

average=0.12 errors/exercise,standard deviation=1.06,mode=0

Intermediate levelaverage=1.52 errors/exercise,standard deviation=1.01,mode=2

average=1.02 errors/exercise,standard deviation=0.98,mode=1

Advanced levelaverage=3,04 errors/exercise,standard deviation=1.04,mode=3

average=2,04 errors/exercise,standard deviation=1.01,mode=2

Table 5.2: Mahjong Lexical Results

Regarding the error rate associated with the collective names and nominal determinants

exercises (see Figure 5.5) shows that Group 1 misses more than Group 2 but this difference be-

comes more critical in the exercise of nominal determinants and collective names (see Figure 5.5).

With this information it is easy to conclude that students feel more difficulties in the collective

names and nominal determinants exercises, and that the “choice of mood in subordinate clause”

are easier (this information is very similar to what students answered in the questionnaire, see

Figure 5.6).

The results of the collective nouns and nominal determinants exercises were very similar.

Regarding the first group, it is noted that on average students miss one exercise in every three

that were submitted and the second group misses less, which is naturally understandable given

46 CHAPTER 5. EVALUATION

Figure 5.5: Error rate by exercise

Figure 5.6: Which exercise was the easiest

the age difference and educational level of the subjects from each group.

In relation to the overall assessment of the system, 50% of the subjects classify it as “very

good”, 17% as “good”, 9% as “acceptable” and the remaining 18% classified it as “bad” or “very

bad” (see Figure 5.7). In this way a global positive appreciation was achieved (67% good or

very good). Group 2 is more critical about the system: only 41% found it very good against

55% from Group 1.

The most critical opinion relates to the high repetition of the infinitive in the “‘mood of

subordinate clauses” exercises. This is due to the fact that most of the stems found in the

CETEMPublico corpus contain the verb in the infinitive, which makes the correct tense for

these exercises also the infinitive. This can lead the users to sense that this exercise is the

47

Figure 5.7: System’s global appreciation

easiest (see Figure 5.6), due to repetition, but also the less interesting one.

One of the ways we can use to help improve this situation, is to create a new condition

during the automatic generation of the exercise that would provide greater variety in the mood

of the stem target verb and thus prevent the exercises selected to users to have always the correct

verb in the infinitive.

Finally, Table 5.3 presents the tests duration. We can verify that Group 1 takes on average

more 5 minutes than the Group 2 to take the test, the Mahjong Lexical being the game that

takes the longest time to complete. Notice that for each exercise in Mahjong 4 pairs are involved,

however in the remaining exercises it is only necessary to select a single answer.

48 CHAPTER 5. EVALUATION

GroupMahjongLexical

Choice ofmood insubordinateclause

CollectiveNames

NominalDetermi-nants

Total

1Average:8m46s;Standard de-viation:0m56s

Average:4m38s;Standard de-viation:1m23s

Average:4m27s;Standard

devia-tion:1m12s;

Average:3m03s;Standard de-viation:1m45s

Average:21m03s;Standard

deviation:1m56s

2Average:7m34s;Standard de-viation:1m45s

Average:3m56s;Standard de-viation:0m46s

Average:3m37s;Standard de-viation:1m09s

Average:2m45s;Standard de-viation:1m24s

Average:15m59s;Standard

deviation:1m46s

Table 5.3: Average test duration

6Conclusion and Future

Work

CALL systems emerged in the context of a growing need for multilingual proficiency in a

globalized world, as a way of providing a fast and appealing tool to learn foreign languages.

This work presented the development of syntatic-semantic module for the REAP.PT system,

able to generate exercises for vocabulary acquisition in a (semi-)automatic way. These exercises

focus on semantic definitions (Mahjong Lexical), on the mood selection for subclause verbs and

on the distributional constraints between Dnom and their heads.

It is important to note that the exercises presented here are especially designed for people

who want to learn Portuguese but that already have a reasonable level of knowledge of the

language, because the generated exercises require the understanding of phrases and definitions

that are difficult concepts for people who have not had contact with the Portuguese.

Regarding the future work, we intend to integrate this system with the DIXI software

(Paulo, Oliveira, Mendes, Figueira, Cassaca, Viana, & Moniz 2008) so that students have the

opportunity to hear the exercises. It is also an objective of this study to make an assessment

of knowledge acquisition by comparing this system with the traditional exercise of textbooks.

Two groups will be used for that - one group uses this system and the others solves written

assignments - at the end we will analyze which group of students obtains better results.

Looking a little more to REAP.PT as a whole, it is important to note that there is still

interesting future work:

• Interface and Portability – The simplicity of the current REAP.PT interface may not

be an attractive point to some students. For that reason, REAP.PT needs to meet the

students’ interests and be adapted to new technologies – devices that the students use in

their everyday life (such as iPhone and iPad) (Correia 2010);

• Games – Carnegie Mellon University started to develop games for inclusion in the REAP

system. The main goal of this proposed future research topic is to make the learning

50 CHAPTER 6. CONCLUSION AND FUTURE WORK

process more attractive and appealing for the students (Correia 2010).

Bibliography

Aherne, A. & C. Vogel (2006). Wordnet enhanced automatic crossword generation. In P. So-

jka, K.-S. Choi, C. Fellbaum, & P. Vossen (Eds.), Proceedings of the Third International

Wordnet Conference, pp. 139 – 145.

Aldabe, I., M. de Lacalle, M. Maritxalar, E. Martinez, & L. Uria. ArikIturri: An Automatic

Question Generator Based on Corpora and NLP Techniques. In M. Ikeda, K. Ashley, & T.-

W. Chan (Eds.), Intelligent Tutoring Systems, Volume 4053 of Lecture Notes in Computer

Science, pp. 584–594. Springer Berlin / Heidelberg.

Anbulagan & A. Botea (2008). Crossword puzzles as a constraint problem. In Proceedings of

the 14th international conference on Principles and Practice of Constraint Programming,

CP ’08, Berlin, Heidelberg, pp. 550–554. Springer-Verlag.

Armand, F. (2001). Learning from expository texts: Effects of the interaction of prior knowl-

edge and text structure on responses to different question types. European Journal of

Psychology of Education 16, pp. 67–86.

Bailey, S. & D. Meurers (2008). Diagnosing meaning errors in short answers to reading com-

prehension questions. In Proceedings of the Third Workshop on Innovative Use of NLP

for Building Educational Applications, EANL’08, Morristown, NJ, USA, pp. 107–115. As-

sociation for Computational Linguistics.

Baptista, J., N. Costa, J. Guerra, M. Zampieri, M. Cabral, & N. Mamede. P-AWL: Academic

word list for portuguese. In T. Pardo, A. Branco, A. Klautau, R. Vieira, & V. Lima (Eds.),

Computational Processing of the Portuguese Language, Volume 6001 of Lecture Notes in

Computer Science, pp. 120 – 123. Springer Berlin / Heidelberg.

Bastos, A., R. Berardi, & R. Silveira (2004). Webduc: Uma proposta de ferramenta de

avaliacao formativa no ambiente TelEduc. In RENOTE Revista Novas Tecnologias na

Educacao, Porto Alegre, v.2, n.1, pp. 161–169.

Carbonell, J. & R. Brown (1988). Anaphora resolution: a multi-strategy approach. In Pro-

51

52 CHAPTER 6. CONCLUSION AND FUTURE WORK

ceedings of the 12th conference on Computational linguistics - Volume 1, Morristown, NJ,

USA, pp. 96–101. Association for Computational Linguistics.

Collins-Thompson, K. & J. Callan (2005). Predicting reading difficulty with statistical lan-

guage models. Journal of the American Society for Information Science and Technol-

ogy 56 (13), pp. 1448–1462.

Correia, R. (2010). MSc Thesis, Automatic Question Generation for REAP.PT Tutoring Sys-

tem, Instituto Superior Tecnico - Universidade Tecnica de Lisboa.

Costa, F. & L. Mendonca (2010). Na Ponta da Lıngua - Lıngua Portuguesa - 6.o Ano. Porto

Editora.

Costa, M. & M. Traca (2010). Passa Palavra - Lıngua Portuguesa - 5.o Ano. Porto Editora.

Costa, R. (2003). Geracao automatica de exercıcios de matematica. Master’s thesis, Departa-

mento de Matematica Pura - Faculdade de Ciencias da Universidade do Porto.

Cristea, P. & R. Tuduce (2005). Automatic generation of exercises for selftesting in adaptive

e-learning systems. In 3rd Workshop on Adaptive and Adaptable Educational Hypermedia

at the AIED’05 conference, (A3EH).

Curto, S. (2010). MSc Thesis, automatic generation of multiple-choice tests , Instituto Supe-

rior Tecnico - Universidade Tecnica de Lisboa.

Dimitrov, A. (2008). Rebuilding WERTi: Providing a platform for second language acquisition

assistance. In Seminar fur Sprachwissenschaft, Universitat Tubingen.

Gamper, J. & J. Knapp (2002). A review of intelligent CALL systems. In J. Gamper &

J. Knapp (Eds.), Computer Assisted Language Learning, Volume 15 of Lecture Notes in

Computer Science, pp. 329–342. Springer-Verlag.

Goto, T., T. Kojiri, T. Watanabe, T. Iwata, & T. Yamada (2010). Automatic generation

system of multiple-choice cloze questions and its evaluation. In International Journal on

Knowledge Management and E-Learning, Volume 2, pp. 210–224.

Hacken, P. T. (2003). Computer-assisted language learning and the revolution in Computa-

tional Linguistics. Journal Linguistik online, Volume 17, pp. 23-39.

Hagege, C., J. Baptista, & N. Mamede (2010). Caracterizacao e processamento de expressoes

temporais em portugues. Linguamatica 2, pp. 63–76.

53

Heilman, M. & M. Eskenazi (2006). Language learning: Challenges for intelligent tutoring

systems. In Proceedings of the Workshop of Intelligent Tutoring Systems for Ill-Defined

Domains. 8th International Conference on Intelligent Tutoring Systems, Volume 8th In-

ternational, pp. 20 – 28.

Holohan, E., M. Melia, D. McMullen, & C. Pahl (2006). The generation of e-learning exercise

problems from subject ontologies. Advanced Learning Technologies, IEEE International

Conference on, pp. 967–969.

Howard, J., M. Kazar, S. Menees, D. Nichols, M. Satyanarayanan, R. Sidebotham, & M. West

(1988). Scale and performance in a distributed file system. ACM Trans. Comput. Syst. 6,

pp. 51–81.

Lisbon European Council (2010). Employment, Economic Reform and Social Cohesion- to-

wards a Europe based on innovation and knowledge. Confederation of British Industry.

Martins, R., G. Nunes, & R. Hasegawa. Curupira: A functional parser for brazilian por-

tuguese. In N. Mamede, I. Trancoso, J. Baptista, & M. das Gracas Volpe Nunes (Eds.),

Computational Processing of the Portuguese Language, Volume 2721 of Lecture Notes in

Computer Science, pp. 195–195. Springer Berlin / Heidelberg.

Marujo, L. (2009). MSc Thesis, REAP em portugues, Instituto Superior Tecnico - Universi-

dade Tecnica de Lisboa.

Meinedo, H., D. Caseiro, J. a. Neto, & I. Trancoso (2003). AUDIMUS : A broadcast news

speech recognition system for the european portuguese language. In N. Mamede, I. Tran-

coso, J. Baptista, & M. das Gracas Volpe Nunes (Eds.), Computational Processing of the

Portuguese Language, Volume 2721 of Lecture Notes in Computer Science, pp. 196–196.

Springer Berlin / Heidelberg.

Meinedo, H. & J. a. Neto (2003). Automatic speech annotation and transcription in a broad-

cast news task. In Proc. ISCA ITRW on Multilingual Spoken Document Retrieval, Hong

Kong.

Moreira, J. E., M. M. Michael, D. da Silva, D. Shiloach, P. Dube, & L. Zhang (2007). Scalabil-

ity of the nutch search engine. In Proceedings of the 21st annual international conference

on Supercomputing, ICS ’07, New York, NY, USA, pp. 3–12. ACM.

Ng, K. (2008). Interactive feedbacks with visualisation and sonification for technology-

54 CHAPTER 6. CONCLUSION AND FUTURE WORK

enhanced learning for music performance. In Proceedings of the 26th annual ACM in-

ternational conference on Design of communication, SIGDOC ’08, New York, NY, USA,

pp. 281–282. ACM.

Okan, Z. Computing laboratory classes as language learning environments. Learning Environ-

ments Research 11, pp. 31–48.

Oliveira, H. G., D. Santos, P. Gomes, & N. Seco (2008). PAPEL: A dictionary-based lexical

ontology for portuguese. In Proceedings of the 8th International Conference on Compu-

tational Processing of the Portuguese Language, PROPOR ’08, Berlin, Heidelberg, pp.

31–40. Springer-Verlag.

O’Malley, J. M., A. U. Chamot, & L. Kupper (1989). Listening Comprehension Strategies in

Second Language Acquisition. Applied Linguistics 10 (4), pp. 418–437.

Paulo, S., L. Oliveira, C. Mendes, L. Figueira, R. Cassaca, C. Viana, & H. Moniz (2008). DIXI

- A Generic Text-to-Speech System for European Portuguese. In A. Teixeira, V. de Lima,

L. de Oliveira, & P. Quaresma (Eds.), Computational Processing of the Portuguese Lan-

guage, Volume 5190 of Lecture Notes in Computer Science, pp. 91–100. Springer Berlin /

Heidelberg.

Pulman, S. & J. Sukkarieh. Automatic short answer marking. In Proceedings of the Second

Workshop on Building Educational Applications Using NLP.

Rost, J. (2005). Software Engineering Theory in Practice. IEEE Softw. 22.

Santos, D. & P. Rocha (2001). Evaluating CETEMPublico, a free resource for Portuguese.

In Proceedings of the 39th Annual Meeting on Association for Computational Linguistics,

ACL ’01, Morristown, NJ, USA, pp. 450–457. Association for Computational Linguistics.

Seixas, M. & C. Castanheira (2010). Aprender em Portugues - Lıngua Portuguesa - 5.o Ano.

978-972-680-565-6. Lisboa Editora.

Siddiqi, R., C. Harrison, & R. Siddiqi (2010). Improving teaching and learning through auto-

mated short-answer marking. IEEE Transactions on Learning Technologies 3, pp. 237–249.

Stark, M. & R. Riesenfeld (1998). Wordnet: An electronic lexical database. In Proceedings of

11th Eurographics Workshop on Rendering. MIT Press.

Stern, H. Practical applications in second language learning. Interchange 6, pp. 57–58.

55

Sukkarieh, J., S. Pulman, & N. Raikes (2003). Auto-marking: using computational linguistics

to score short, free text responses. paper presented at the 29th annual conference. In of

the International Association for Educational Assessment (IAEA).

Tomas, A. & J. Leal (2003). A CLP-Based Tool for Computer Aided Generation and Solving

of Maths Exercises. In V. Dahl & P. Wadler (Eds.), Practical Aspects of Declarative Lan-

guages, Volume 2562 of Lecture Notes in Computer Science, pp. 223–240. Springer Berlin

/ Heidelberg.

Tomas, A., J. Leal, & M. Domingues. A web application for mathematics education. In

H. Leung, F. Li, R. Lau, & Q. Li (Eds.), Advances in Web Based Learning - ICWL

2007, Volume 4823 of Lecture Notes in Computer Science, pp. 380–391. Springer Berlin /

Heidelberg.

56 CHAPTER 6. CONCLUSION AND FUTURE WORK

A5th Grade P{AWL

abandonar benefıcio comprador debate esquema

academia bolsa compreender decidir essencia

acumulacao canal computador declinar estabelecimento

acumular canalizado comunicar deduzir estimar

adaptar canalizar concluir definido estudar

adulto capaz confianca deixar estudo

afectivamente capıtulo confiar depressao eticamente

afixar caracterıstico conformar deprimir eventualidade

ajuda categoria conforme desafio expansao

ajudar chamado consagrar descobrir exploracao

ajuste ciclo conseguir desmotivar exteriorizar

anual citacao constantemente detector extrair

apoio clareza consulta dominar facil

area colapso consultar empresa finalmente

assegurar colega contacto encontrar fısica

assuncao colocar contratar enorme forcar

assunto comecar convencido equivaler formato

atribuicao comeco conversar errar formular

aumentar compor converso esclarecer funcionamento

avaliar compra corporacao esclarecimento fundacao

fundar labor paragrafo rejeitar texto

fundo legislacao parcialmente renda total

grafico libertar participar repartir trabalho

heranca licenciado participativo requisitar transporte

ignorante localizar perıodo resolver unicamente

ilegalidade maduro perseguicao responder util

imaginar manual perseguir resposta variar

imaginativo mapa persistente resumir vestıgio

imaturo marcar polemica retencao visivelmente

Table A.1: 5th Grade P–AWL (1st Part)

58 APPENDIX A. 5TH GRADE P–AWL

incessante margem posar revelacao inconformado

maximo preciso rever indiscreto medico

princıpio rota individualizar mental problema

salientar inevitavel meta procurar seccao

informacao ministro profissao seguranca informatico

modalidade profissional sexo insignificante monitor

projectar significado inspeccionar montar propriedade

somar instrutivo negar racional subsıdio

instrutor negativa rascunho sucesso inteligente

numero reaccao suposicao intensamente obter

realizar tabela intervalo ocupacao receita

tarefa inutil orientar recuperar tecnica

inutilmente origem recurso tecnico invalido

palestra rede tematica juntar par

regime terreno

Table A.2: 5th Grade P–AWL (2nd Part)

BNumber of exercises for

each nominal determinant

B.1 Collective names

acervo – 51 agremiacao – 8 album – 579 alcateia – 2 antologia – 166

armada – 5 arquipelago – 15 arsenal – 57 baixela – 0 banda – 470

bando – 192 batalhao – 169 bateria – 45 bosque – 6 bracado – 2

bracada – 2 brigada – 188 buque – 0 bouquet – 0 cacho – 12

cafila – 3 caravana – 59 cardume – 8 catalogo – 136 caterva – 12

charanga – 1 choldra – 0 chorrilho – 34 chusma – 12 claque – 1

colecao – 0 coleccao – 0 colegio – 54 coletanea – 0 colectanea – 0

colmeia – 6 colonia – 45 colonia – 45 comboio – 132 comissa – 588

comitiva – 67 companhia – 214comunidade –257

constelacao – 49 cordilheira – 2

corja – 2 coro – 226 elenco – 181 enxame – 16 enxoval – 0

equipa – 2877 esquadra – 31 esquadrilha – 16 exercito – 160 fardo – 17

feixe – 44 floresta – 36 fornada – 31 frota – 123 girandola – 9

horda – 15 junta – 9 juri – 109 legiao – 121 leva – 0

magote – 18 manada – 28 mata – 16 matilha – 3 mirıade – 47

molho – 45 multidao – 374 ninhada – 3 nuvem – 94 orquestra – 90

pilha – 52 plantel – 60 plateia – 0 pomar – 5 praga – 4

prole – 5 quadrilha – 16 rajada – 36 ramalhete – 10 ramo – 137

rancho – 10 rebanho – 27 recova – 0 recua – 0 renque – 2

restia – 25 revoada – 0 saraivada – 10 seleta – 0 trem – 1

tripulacao – 51 trouxa – 2 turma – 60 vara – 10

Table B.1: Collective names

60 APPENDIX B. NUMBER OF EXERCISES FOR EACH NOMINAL DETERMINANT

B.2 Nominal Determinants

alforge – 3 alguidar – 4 anfora – 1 arca – 9 bacia – 9

balde – 33 bilha – 3 bolsa – 132 bornal – 0 botija – 0

bureta – 0 cabaca – 0 cabaz – 24 cacarola – 2 caixa – 245

caixote – 20 caldeira – 4 caldeirao – 20 calice – 31 canastra – 0

caneca – 8 capsula – 19 carga – 147 carteira – 88 cartucho – 6

cesta – 5 cesto – 43 concha – 4 contentor – 22 copo – 139

cuba – 0 escudela – 0 estojo – 6 frasco – 33 gamela – 1

garrafa – 285 garrafao – 3 gaveta – 10 jarra – 10 jarro – 5

lata – 73 maco – 47 mala – 13 maleta – 0 malga – 6

marmita – 0 meada – 0 mesa – 78 molheira – 0 moringa – 0

pacote – 704 palete – 1 panela – 10 pasta – 34 pedaco – 345

pipa – 9 pires – 2 pitada – 33 prato – 123 recipiente – 10

restea – 20 saca – 5 saco – 105 sacola – 3 selha – 0

taca – 42 tacho – 11 tanque – 30 termo – 37 terrina – 2

tigela – 3 tina – 1 tonel – 3 urna – 5 vasilha – 3

vaso – 17 xıcara – 0chavena de cafe –0

chavena de cha –0

colher de cafe – 4

colher de cha – 14colher de so-bremesa – 6

colher de sopa –26

data – 178 imensidade – 13

montanha – 55 monte – 189 parte – 4768percentagem –577

porcao – 78

porradao – 0sem-numero –148

Table B.2: Nominal determinants

CInitial Form

Figure C.1: Initial Form

62 APPENDIX C. INITIAL FORM

DQuestionnaire

Figure D.1: Questionnaire - 1st Part

64 APPENDIX D. QUESTIONNAIRE

Figure D.2: Questionnaire - 2nd Part