harnessing disagreement for event semantics

27
Harnessing Disagreement for Event Semantics Lora Aroyo Chris Welty

Upload: lora-aroyo

Post on 06-May-2015

699 views

Category:

Technology


0 download

DESCRIPTION

http://ceur-ws.org/Vol-902/paper_4.pdf The focus of this paper is on how events can be detected & extracted from natural language text, and how those are represented for use on the semantic web. We draw an inspiration from the similarity between crowdsourcing approaches for tagging and text annotation task for ground truth of events. Thus, we propose a novel approach that harnesses the disagreement between the human annotators by defining a framework to capture and analyze the nature of the disagreement. We expect two novel results from this approach. On the one hand, achieving a new way of measuring ground truth (performance), and on the other hand identifying a new set of semantic features for learning in event extraction.

TRANSCRIPT

Page 1: Harnessing Disagreement for Event Semantics

Harnessing Disagreement for Event Semantics

Lora Aroyo Chris Welty

Page 2: Harnessing Disagreement for Event Semantics

Objects vs. Events

events perdure = their parts exist at different time points objects endure = they have all their parts at all points in time

objects are wholly present at any point in time, events unfold over time

Flickr: vanilllaph

Page 3: Harnessing Disagreement for Event Semantics

Event Extraction in NLP

•  Find Events mentioned in text •  Type them (communication, bombing, …) •  Find the role fillers (location, date,

participants, …)

•  State of the art is extremely low: .10 F

Page 4: Harnessing Disagreement for Event Semantics

Measuring •  Define the task for human annotators •  Annotate by multiple people •  Measure agreement •  While k<.6 refine definition and repeat

... for events this process is long, disagreement

is high, agreement becomes forced

Page 5: Harnessing Disagreement for Event Semantics

Position

Human disagreement about events are part of their semantics

Flickr: elkabong

Page 6: Harnessing Disagreement for Event Semantics

Events are Vague Humans have no clear notion of what

events are

Page 7: Harnessing Disagreement for Event Semantics

“event is a significant "happening" or gathering of people. I would define a "happening" as an event if the group of people gathered were united in one common goal.”

We Asked the Crowd What an EVENT is Flickr: massimo vitali

Page 8: Harnessing Disagreement for Event Semantics

We Asked the Crowd What an EVENT is

Event is a happening, which can be scheduled or unscheduled. An earthquake or fire happens (unscheduled). A wedding or birthday party (scheduled). It is an occasion that is unusual and tends to be memorable.

Page 9: Harnessing Disagreement for Event Semantics

“An event would be any occurrence where physical action has taken place. It may be a single, momentary instance (I sneezed), or it may span a period of time (the festival ran for four hours). An event may also be made up of a number of smaller events, such as a day at school is an event, but each individual class is also an event itself. Basically an event must have a physical action over any delimited time span.”

We Asked the Crowd What an EVENT is

Page 10: Harnessing Disagreement for Event Semantics

“A planned public or social get together or occasion.”

“an event is an incident that's very important or monumental”

“An event is something occurring at a specific time and/or date to celebrate or recognize a particular occurrence.”

“a location where something like a function is held. you could tell if something is an event if there people gathering for a purpose.”

“Event can refer to many things such as: An observable occurrence, phenomenon or an extraordinary occurrence.”

We Asked the Crowd What an EVENT is

Page 11: Harnessing Disagreement for Event Semantics

What do Experts think an EVENT is?

“an event is the exemplification of a property by a substance at a given time” Jaegwon Kim, 1966

“events are changes that physical objects undergo” Lawrence Lombard, 1981

“events are properties of spatiotemporal regions”, David Lewis, 1986

under30ceo.com

Page 12: Harnessing Disagreement for Event Semantics

What do Experts think an EVENT is?

“an event is the exemplification of a property by a substance at a given time” Jaegwon Kim, 1966

“events are changes that physical objects undergo” Lawrence Lombard, 1981

“events are properties of spatiotemporal regions”, David Lewis, 1986

under30ceo.com

nothing everything

Page 13: Harnessing Disagreement for Event Semantics

Why is event semantics hard?

Page 14: Harnessing Disagreement for Event Semantics

the World is Open

1. events have multiple dimensions 2. each dimension has levels of granularity

3. people have different views on both

all this leads to very complex semantics

Page 15: Harnessing Disagreement for Event Semantics

and our goal is ...

1. not to enforce agreement 2. to capture different view points

3. to teach machines to reason in the disagreement space

Flickr: elkabong

Page 16: Harnessing Disagreement for Event Semantics

Hypothesis Artificially restricting humans does not help machines to learn.

Machines will learn from diversity

Flickr: elkabong

Page 17: Harnessing Disagreement for Event Semantics

What do People Disagree on?

are sub-events always mere parts? are “mentions” meaningful for events?

are events coreferential across documents? (e.g. perspectives, observations)

Page 18: Harnessing Disagreement for Event Semantics

“the bombing targeted a housing development in Baghdad, killing 3 and injuring 13”

indistinguishable by people, confusable: is bombing part of killing, or killing part of bombing?

What about targeting?

“merelogically extensional” (i.e arbitrary): container bursting into fragments as a result of explosion

some events don’t exist:

an action by military forces prevented the bombing.

Page 19: Harnessing Disagreement for Event Semantics

Disagreement Framework

•  ontology: disagreements on the basic status of events themselves as referents of linguistic utterances, e.g. are people events or do events exist at all.

•  granularity: disagreements that result from issues of granularity, e.g. the location being a country, region, or city, the time being a day, week, month, etc.

•  interpretation: disagreements that result from (non-granular) ambiguity, differences in perspective, or error in interpreting an expression, e.g. classifying a person as a terrorist/hero, ”October Revolution” took place in September.

Page 20: Harnessing Disagreement for Event Semantics

Granularity Disagreement

•  spatial, temporal, participants

•  compositional, classificational

Page 21: Harnessing Disagreement for Event Semantics
Page 22: Harnessing Disagreement for Event Semantics

Event Participants Disagreement

Prime minister Benjamin

Netanyahu

Benjamin Netanyahu

Israeli Prime minister

Cabinet

Benjamin Netanyahu’s

Cabinet

Israeli Cabinet

his Cabinet

Israeli Government

{TOLD}

50%

35%

15%

10%

15%

5%

45%

15%

Page 23: Harnessing Disagreement for Event Semantics

Temporal Disagreement

Spring 1998

March 1, 1998

March 1998

Sunday Prime minister

Benjamin Netanyahu

Benjamin Netanyahu

Israeli Prime minister

{TOLD}

50%

35%

15%

25%

15%

50%

5%

Page 24: Harnessing Disagreement for Event Semantics

Spatial Disagreement

Lebanon

Israel Southern Lebanon

Israel's Northern Frontier

Middle East

{WILLING TO WITHDRAW}

35%

45%

10%

30%

65%

Page 25: Harnessing Disagreement for Event Semantics

Approach Principles

1.  tolerate, capture & exploit disagreement 2.  understand the range of disagreements

by creating a space of possibilities with frequencies & similarities

3.  score machine output based on where it falls in this space

4.  adaptable to new annotation tasks

Flickr: auroille

Page 26: Harnessing Disagreement for Event Semantics

Position

Human disagreement about events are part of their semantics

Flickr: elkabong

Page 27: Harnessing Disagreement for Event Semantics

Conclusion Artificially restricting humans does not help machines to learn.

Machines will learn from diversity

Flickr: elkabong