Horizons for Information Societies
Seminar #10

The 10^th seminar dealt with data processing. Dr Evans () presented his group's activities on data retrieval and his research on multilingual opinion analysis at .

Setting

Slides of the presentations:

Dr EVANS David:
Dr EVANS David:

Date: 26 February 2008 (15:30-17:00)
Location: , Tokyo, Japan
Language: English
Registration fees: None
Attendees: 8 persons
Organization: Dr ,

Talks

Overview of the NTCIR Information Access Evaluation Workshop
by Dr

Abstract: The NTCIR Workshop is a series of evaluation workshops designed to enhance research in Information Access technologies including information retrieval, question answering, text summarization, extraction, etc. The workshops have been held every year and half since 1999, and the seventh workshop will be held 16-19 December 2008 in 東京 (Tokyo, Japan). Past workshops focused on improving cross-language information retrieval, creating large corpora of Japanese, Chinese, English, and Korean documents for Information Retrieval research. As information retrieval has become a more common concept, epitomized by daily use of Google.com, our workshop moved on to more challenging areas of information access evaluation. In the 2008 NTCIR-7 Workshop, we will have tracks in three different clusters:

Advanced Cross-Lingual Information Access

Complex Cross-Lingual Question and Answering, where systems accept questions in English, and return answers in either Japanese or Chinese
Information Retrieval for Question Answering, which focuses on the information retrieval aspect of question answering

User Generated Contents

Multilingual Opinion Analysis Task, focusing on sentiment analysis in text
Cross-Lingual Information Retrieval for Blogs, which extends CLIR evaluations into a new genre of user-generated blogs

Focused Domains

Patent Translation, which investigates automatic Japanese-English translation using a new two-million sentence pair aligned Japanese-English corpus
Patent Mining, which looks at categorization of Patent Applications to aid in retrieval, and technological term extraction for trend identification

I will present and overview past NTCIR Evaluation Workshops as well as the current evaluation tasks for 2008.

Speaker: David Kirk Evans received a Masters of Science and a Doctorate of Philosophy in Computer Science from Columbia University in 2002 and 2005, respectively. There, he studied in the Natural Language Processing group advised by Professor Judith Klavans, and worked closely with Professor Kathy McKeown on multilingual, multidocument summarization. His interests include multidocument summarization with a focus on multilingual processing, machine translation, and pedagogical applications of computational linguistics. He moved to Japan in March 2006 and worked as a Post-Doc at with in cross-lingual information retrieval, summarization, and sentiment analysis.

Website: <http://research.nii.ac.jp/ntcir/>

Download

Multilingual Opinion Analysis at the NTCIR Evaluation
by Dr

Abstract: I will introduce the NTCIR6 Opinion Analysis Pilot, information about the Chinese, Japanese, and English data, plans for future opinion analysis tasks, and give overviews of the systems submitted by the task participants. The task is a sentence-level opinion identification and polarity detection task run over data from a comparable corpus in Chinese, English, and Japanese. We manually annotated documents for this task in each language, producing a multilingual opinion analysis data set over comparable data. I will describe results from the evaluation, plans for the NTCIR7 Multilingual Opinion Analysis Task, and recent work I have done analyzing the automatic identification of opinion holders from text.