Thematic Sessions

When you submit a paper, you will be asked to indicate whether it is being submitted to the General Sessions or to one of the Thematic Sessions. The following gives the list of the Thematic Session titles for ACL-2000, each with a short description of the topics to be covered in that theme so that you can determine which theme you want to submit your paper to. Note that the topic lists shown here are intended to be indicative rather than exhaustive; follow the links to the theme web pages to find more information about the intent of the themes.

T1: NLP and Open-Domain Question Answering from Text

Chair: Sanda Harabagiu
Motivation: The recent explosion of information available on the World Wide Web determines question answering (Q/A) to be a compelling framework for finding information that closely matches user needs. Several current NLP-based technologies are able to provide the framework that approximates the complex problem of answering questions from large collections of texts.
List of topics: Topics of interest include, but are not limited to the following:

analysis of natural language questions
lexical resources in natural language Q/A
NLP for internet Q/A text mining for natural language Q/A

role of information extraction (IE) in natural language Q/A
indexing methods for natural language Q/A
empirical methods for answer detection in natural language Q/A
abductive techniques for justification of answer correctness
coreference resolution for natural language Q/A tasks

http://renoir.seas.smu.edu/~sanda/acl2000-topic.html

T2: Machine learning and statistical NLP for dialogue

Chairs: Antal van den Bosch, Emiel Krahmer, Maria Wolters
Motivation: The theme addresses the question to what extent the traditional approach to developing dialogue systems can be supplemented by automatic data-driven techniques from ML and statistical NLP. The crucial advantage of such techniques is that they do not require the development of a formal model beforehand, but instead extract relevant knowledge directly from annotated data. The aim of this theme session is to bring representatives of linguistically-oriented and computationally-oriented groups together, to foster exchange between those who provide the theoretial tools for successful application of ML and statistical NLP techniques, those who work on discourse and dialogue theory, and those who build actual dialogue systems.
Topics:

Suitability of dialogue system components for incorporation of Machine Learning (ML) and statistical modeling
Application and applicability of state-of-the-art Machine Learning and statistical NLP techniques to (components of) dialogue systems (speech recognition and synthesis, dialogue management, language interpretation and generation, coreference processing, error detection techniques etc.)
Consequences of incorporating data-driven modules for dialogue system architecture
Corpus collection and annotation
Evaluation of dialogue system components based on ML/statistical NLP

http://ilk.kub.nl/acl00-theme/

T3: Text Summarization

Chair: Inderjeet Mani
Motivation: Automatic summarization aims at providing a condensed representation of the content of an information source in a manner sensitive to the needs of the user and task. As we emerge into the 21st Century, the massive information universes that lie ahead make some form of text summarization indispensable. While the field continues to progress, there are also many problems in summarization per se, as well as in areas of NL understanding and NL generation, that need to be addressed before the promises of automatic text summarization can be fully realized.
Topics: Multilingual, multidocument, and multimedia summarization, NL generation for summarization, statistical models, narrative techniques, topic identification and concept fusion for summarization, evaluation methods, development and exploitation of summarization resources, practical applications.
http://www.geocities.com/Athens/Forum/1373/acl-2000-summarization-theme.html

T4: Theoretical and Technical Approaches for Asian Language Processing --Similarities and Differences among Languages
- Chairs: Hitoshi Isahara, Rajeev Sangal, Ming Zhou
- Motivation: Research on natural language processing (NLP) for Asian languages has begun to flourish especially since electronic data has become available in some Asian languages. Theoretically, the techniques adopted for one language are applicable for any language, however, there still might be some differences among languages and that makes it difficult to transfer the advances in one language to other languages. This theme session will provide a forum for researchers in the area of NLP who are interested in advancing the state of development of NLP techniques for Asian languages.
- Topics:
  - Methodological Aspects of Asian Language Processing: Corpus-based, Learning and Statistical Approach, Data Sparseness, Evaluation, and (Semi-)Automatic Annotation
  - Theoretical Aspects of Asian Language Processing: Polite Expressions, Free Word Order Languages, Combining Corpus-based and Rule-based Approaches, Computational Models Including Aspects of Syntax and Semantics, and Lexical Semantics
  - Application: Information Retrieval for Mono/Multi Asian Languages, Classification of Texts using Statistical Means, and Machine Translation among/from/to Asian Languages
  - Linguistic Resources for Asian Language Processing Multilingual/Monolingual/Learners Corpora, Machine Readable (Mono/Multi- lingual) Dictionaries, Extraction of Lexical Information from Corpora, and Development of Benchmark Data Sets for NLP
- http://www-karc.crl.go.jp/ips/members/isahara/ACL2000TS/index.html

Last modified: Tue Jan 18 10:12:11 EST 2000