socio semantic networks of research publications in learning analytics community
TRANSCRIPT
page 1
Socio-semantic Networks of Research Publications in the Learning Analytics CommunitySoude Fazeli, PhD candidateDr. Hendrik DrachslerProf. Dr. Peter Sloep
page 2
Agenda
1. Introduction
2. Motivation
3. Data processing
4. Network visualization
5. Discussion and conclusion
page 3
3
1. Introduction
• Presenting visualization of the authors and papers network• Carrying out a deeper analysis of the generated networks • The main aim
• To use such a graph of authors and papers to recommend similar items to a target user
24
42
10
51
119
27
0
20
40
60
80
100
120
140
LAK 2011 LAK 2012 JETS2012
Papers
Authors
page 4
2. Motivation
• A list of recommended authors and papers
• To plan the conference participation more efficiently and effectively
• Awareness support for researchers (Reinhardt et al., 2012; Fisichella et al.,
2010; Ochoa et al., 2009; Henry et al., 2009)
• Scientific recommender systems (Huang et al., 2002; Wang & Blei, 2010)
• Such a priority list
• Support the awareness of the attendees
• Empower the network of like-minded authors in the attendees’ particular research focus
page 5
• RQ1. How are the authors connected and which authors share more connections and are more central in terms of sharing commonalities with the others?
• RQ2. How are the papers connected to each other in terms of similarity?
page 6
1. Finding patterns of similarity between authors and papers2. Visualizing networks of the LAK authors and papers
page 7
3. Data processing3.1. Tag clouds
• The TF-IDF algorithm• Weighted list of the most commonly used terms in research
articles
• Using default algorithm provided by Mahout on the text files extracted from the RDF files
• Removing the stop words • Setting the configuration variables within Mahout to 90%
• Outcome• A so-called dictionary of all the terms in the LAK dataset
• A binary sequence file including the TF-IDF weighted vectors
page 8
3. Data processing3.2. Computing similarity
• Using T-index algorithm (Fazeli et al., 2010)
• Collaborative filtering recommender algorithm
• Generates a graph of users• The nodes are users and the edges show the relationship between
users • Originally makes recommendations based on the ratings data of users
• Extending the T-index algorithm
• Process tags and keywords extracted from the linked data
• Using Jena APIs to • Process RDF files
• Handle Ontology Web Language (OWL) files describing the generated graph of authors and papers
page 9
4. Network visualization4.1. Author network (made by Welkin)
page 10
4. Network visualization4.1. Author network
The degree centrality of the top ten central authors
121
9285
7164
5550 49
45 44
96
57 5546 45
36 35
2217 16
0
20
40
60
80
100
120
140
u1 u2 u3 u4 u5 u6 u7 u8 u9 u10
Then first ten central authors
inde
gree
n=10
n=5
page 11
4. Network visualization4.2. Paper network
page 12
4. Network visualization4.2. Paper network
3430
24 23 21 20 19 18 17 16
58
5147 46
42
34 33 3127 26
0
10
20
30
40
50
60
70
p1 p2 p3 p4 p5 p6 p7 p8 p9 p10
n=5
n=10
The degree centrality of Top ten papers
page 13
• RQ1. How are the authors connected and which authors share more connections and are more central in terms of sharing commonalities with the others?
• RQ2. How are the papers connected to each other in terms of similarity?
page 14
5. Discussion and conclusion5.1. RQ1
The first ten central authors
Author DegreeHendrik Drachsler 116
Kon Shing Kenneth Chung 87
Wolfgang Greller 80
Javier Melenchon 66
Brandon White 59
Vania Dimitrova 50
Erik Duval 45
Rebecca Ferguson 44
Anna Lea Dyckhoff 40
Simon Buckingham Shum 39
page 15
• RQ1. How are the authors connected and which authors share more connections and are more central in terms of sharing commonalities with the others?
• RQ2. How are the papers connected to each other in terms of similarity?
page 16
5. Discussion and conclusion5.2. RQ2
Paper Authors
Learning Dispositions and Transferable Competencies: Pedagogy, Modelling and Learning Analytics
Simon Buckingham-Shum, Ruth Deakin Crick
The Pulse of Learning Analytics Understandings and Expectations from the Stakeholders
Hendrik Drachsler, Wolfgang Greller
Social Learning Analytics: Five Approaches Rebecca Ferguson, Simon Buckingham-Shum
Multi-mediated Community Structure in a Socio-Technical Network
Dan Suthers, Kar Hai Chu
Modelling Learning & Performance: A Social Networks Perspective
Walter Christian Paredes, Kon Shing Kenneth Chung
Teaching Analytics: A Clustering and Triangulation Study of Digital Library User Data
Beijie Xu, Mimi M Recker
Monitoring Student Progress Through Their Written "Point of Originality"
Johann Ari Larusson, Brandon White
Learning Designs and Learning Analytics Lori Lockyer, Shane Dawson
A Multidimensional Analysis Tool for Visualizing Online Interactions
Eunchul Lee, M'hammed Abdous
Using computational methods to discover student science conceptions in interview data
Bruce Sherin
The top ten central papers
page 17
5. Discussion and conclusion
• The central authors and top ten papers will appear more often in the top recommendation list
• Although most of the central authors also appear in top ten papers’ list, the order is not the same
• Some authors has more than one paper
• Not each and every one of the authors’ papers individually has the highest similarity to the other papers
page 18
Soude FazeliPhD candidateOpen University of the NetherlandsCentre for Learning Sciences and Technologies(CELSTEC) PO-Box 29606401 DL Heerlen, The Netherlandsemail: [email protected]