Current Trend in Information Extraction

Hauptseminar: Computerlinguistik, 2. Studienabschnitt

Leitung: Günter Neumann
Ort: Geb. C7 1, Besprechungsraum U.15
Zeit: Do 11 - 13 Uhr
Beginn: 2nd November
Geeignet für: B.Sc./M.Sc.

Abstract

In this course we will consider new methods and strategies for Information Extraction (IE) with a focus on Machine Learning and Evaluation. We will start by presenting a brief overview of IE. We will then present recent Machine Learning approaches for Named Entity recognition, relation extraction, and event extraction. This will also cover their use in specific applications, e.g., bioinformatics. Beside the presentation of these IE methods, we will also present major trends in evaluation of IE systems, e.g., Automatic Content Extraction (ACE), BioCreative challenge.

Seminar language: English

Certificates
talk or talk + paper + participation (no less than 8 sessions).

Stellung im Studienplan
als Hauptseminar im B.Sc.: Regelstudienzeit 6. Semester;
as seminar in M.Sc. Programm: Standard time slot 2nd semester. specialization area: LT

Credit points:
M.Sc.: talk only 4 LP (5LP); talk + paper 7 LP (9LP); B.Sc.: talk + paper 7LP (9LP)

Schedule (Topics)

02.11.2006	Initial	Theme discussion	Speaker	Presentation

09.11.2006	Session 1	IE overview	Günter Neumann	IE-overview.pdf
16.11.2006	No Session	No Topic
23..11.2006	Session 2	The Automatic Content Extraction (ACE) Program	Akira Kakinohana	ACE.pdf
30.11.2006	Session 3	BioCreative Challenge - Assessment for IE in biology	Jana Besser Caroline Maginot	BioCreative-1.pdf BioCreative2.pdf
06.12.2006	Session 4	Shallow Parsing for biomedical text	Philip Kellner Stefan Kazalski	postponed to 15.02.2007 ClosedClassParser.pdf
14.12.2006	No Session	No Topic
21.12.2006	Session 5	Named Entity recognition for biomedical text	Achmad Yani Yu Fu	BioNER.pdf HMMBioNER.pdf
11.01.2007	Session 6	Complex Relation Extraction with Applications to Biomedical IE	Michaela Regneri Niko Felger	FindingClieques.pdf ExtractComplexRelations.pdf
18.01.2007	Session 7	Preemptive IE	Andrea Heyl	OnDemand&PreemptiveIE.pdf
25.01.2007	Session 8	Expressing Implicit Semantic Relations without Supervision	Yu Chen Lu Zhang	SemanticSimilarity.pdf UnsupervisedRelSim.pdf
01.02.2007	Session 9	Paraphrase Extraction	Teresa Herrmann Anna Mündelein	ParaExtrRTE.pdf UnsupervisedParaExtr.pdf
08.02.2007	Session 10	Machine Learning for Question Answering	Olga Cander Michael Wirth	AnswerPrediction.pdf
15.02.2007	Session 11	Shallow Parsing for biomedical text Final meeting and discussion	Philip Kellner ALL

Paper Report

For the report writing, we will use a standard conference style which is available for Latex and MS-word. The ZIP file contains corresponding versions for the format instructions. Important:

The paper lenght is restricted to eight (8) pages.
Deadline for submission is the will be announced.
Submission should be electronically made to my email address!

References

Session 1: IE overview

Session 2: The Automatic Content Extraction (ACE) Program

Main links of AC:

ACE annotation tools:

http://projects.ldc.upenn.edu/ace/tools/

Session 3: BioCreative Challenge - Assessment for IE in biology

Main link to BioCreative:

http://biocreative.sourceforge.net/

Summary of BioCreative 1 challenge:

Overview of BioCreAtIvE: critical assessment of information extraction for biology. Lynette Hirschman, Alexander Yeh, Christian Blaschke, Alfonso Valencia BMC Bioinformatics 2005, 6(Suppl 1):S1 (24 May 2005)
[PDF]
BioCreAtIvE Task 1A: gene mention finding evaluation. Alexander Yeh, Alexander Morgan, Marc Colosimo, Lynette Hirschman BMC Bioinformatics 2005, 6(Suppl 1):S2 (24 May 2005)
[PDF]
Evaluation of BioCreAtIvE assessment of task 2. Christian Blaschke, Eduardo Andres Leon, Martin Krallinger, Alfonso Valencia BMC Bioinformatics 2005, 6(Suppl 1):S16 (24 May 2005)
[PDF]

Session 4: Shallow Parsing for biomedical text

Main papers:

A Shallow Parser Based on Closed-Class Words to Capture Relations in Biomedical Text. Gondy Leroy, Hsinchun Chen, Jesse D. Martinez (2003)
[PDF]
Extensible Shallow Parsing for Semantic Nets. J. Connell (2001)
[PDF]
A language--independent shallow--parser compiler. A. Kinyon (2001)
[PDF]

Session 5: Named Entity recognition for biomedical text

Main links:

Session 6: Complex Relation Extraction with Applications to Biomedical IE

Main paper and additional resources:

Simple Algorithms for Complex Relation Extraction with Applications to Biomedical IE. R. McDonald, F. Pereira, S. Kulick, S. Winters, Y. Jin and P. White
[PDF]
MALLET - Advanced Machine Learning for Language - a Java based ML Tool box
Algorithm 457: finding all cliques of an undirected graph. C. Bron and J. Kerbisch (1973)
[PDF]

Session 7: Preemptive IE

Main papers:

Preemptive Information Extraction using Unrestricted Relation Discovery. Yusuke Shinyama, Satoshi Sekine (2006)
[PDF]
On-Demand Information Extraction. Satoshi Sekine (2006)
[PDF]

Session 8: Expressing Implicit Semantic Relations without Supervision

Main papers:

Expressing implicit semantic relations without supervision. Turney, P.D. (2006)
[PDF]
Similarity of semantic relations. Turney, P.D. (2006)
[PDF]

Session 9: Paraphrase Extraction

Main papers:

Unsupervised Paraphrase Acquisition via Relation Discovery. Takaaki Hasegawa (2005)
[PDF]
Paraphrase Substitution for Recognizing Textual Entailment. Wauter Bosma and Chris Callison-Burch (2006)
[PDF]

Session 10: Machine Learning for Question Answering

Main papers:

Language Independent Answer Prediction from the Web. A. Figueroa and G. Neumann (2006)
[PDF]
Instance-Based Question Answering: A Data Driven Approach. L. Lita and J. Carbonell (2004)
[PDF]

Links

My ESSLLI 2004 course in Information Extraction

ACE - Automatic Content Extraction

PBioCreative challenge evaluation

Evaluating Machine Learning for Information Extraction