Skip to main content

Yasushi Ogawa

Follower

1

Following

1

Public Views

Address: Yokohama, Kanagawa, Japan

less

Katherine Butler Schofield

King's College London

David Goldstein

University of California, Los Angeles

Norwegian University of Science and Technology

Università di Bologna

Antonis K . Petrides

Open University of Cyprus

Yannis Hamilakis

Brown University

Francesco Camia

Università degli Studi "La Sapienza" di Roma

University of Toronto

Javier Marín-López

Universidad de Jaén

Princeton University

Interests

Uploads

Papers by Yasushi Ogawa

Supervised Synonym Acquisition Using Distributional Features and Syntactic Patterns

Journal of Natural Language Processing, 2009

Distributional similarity has been widely used to capture the semantic relatedness of words in ma... more Distributional similarity has been widely used to capture the semantic relatedness of words in many NLP tasks. However, parameters such as similarity measures must be manually tuned to make distributional similarity work effectively. To address this problem, we propose a novel approach to synonym identification based on supervised learning and distributional features, which correspond to the commonality of individual context types shared by word pairs. This approach also enables the integration with pattern-based features. In our experiment, we have built and compared eight synonym classifiers, and showed a drastic performance increase of over 60% on F-1 measure, compared to the conventional similarity-based classification. Distributional features that we have proposed are better in classifying synonyms than the conventional common features, while the pattern-based features have appeared almost redundant.

Optimizing query evaluation in n-gram indexing

Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval

In n-gram indexing widely used in Asian languages, because a long query word is divided into more... more In n-gram indexing widely used in Asian languages, because a long query word is divided into more than one n-gram, proximity constraints among ngrams need to be checked. And this checking degrades the retrieval speed. This paper proposes two optimizations of query evaluation: one selects a combination of n-grams whose processing cost is the minimum among all the possible combinations, and the other uses additional n-grams in finding potential documents whose proximity constraints need to be checked.

Simple word strings as compound keywords

Proceedings of the 16th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '93, 1993

This paper describes a new indexing method for Japanese text databases using the simple keyword s... more This paper describes a new indexing method for Japanese text databases using the simple keyword string, in which a compound word is treated as a string of simple words, which are the smallest units in Japanese grammar which still maintain their meanings. This method allows retrieved texts to be ranked, according to the similarity of their meaning to the query, without using a control vocabulary or thesaurus. This paper also introduces the keyword feature, which describes the syntactic and semantic characteristics of a word, and results in more precise keyword extraction and text retrieval as well as simple dictionary maintenance.

A New Indexing and Text Ranking Method for Japanese Text Databases Using Simple-Word Compounds as Keywords

Database Systems for Advanced Applications '93, 1993

This paper describes a new indexing method for Japanese text databases using the simpie keyword s... more This paper describes a new indexing method for Japanese text databases using the simpie keyword string. A compound word is treated as a string of simple words, which are the smallest units in Japanese grammar which still maintain their meanings. As a result, retrieved texts can be ranked, according to the similarity of their meaning and the query, without using a control vocabulary or thesaurus. For automatic indexing, the newly introduced keywordfeature, which describes the syntactic and semantic characteristics of a word, enables a precise keyword assignment and simplifies dictionary maintenance. Lexical and Syntactic Parsing: Since Japanese text contains no obvious space between (simple) words, the text must be lexically and syntactically parsed to identify simple words. Our system utilizes the Japanese parser developed at our Research CenterD]. A part of speech and keyword features, which are explained later, are also assigned to each simple word. Keyword Candidate Extraction: Based on the part of speech, simple words are extracted from the parsed text as keyword candidates.

Bucket Spreading Parallel Hash: a New, Robust, Parallel Hash Join Method for Data Skew In the Super Database Computer (SDC)

… of the sixteenth international conference on …, 1990

The Super Database Computer (SDC) is a high-performance relational database server for a join-int... more

RICOH at TREC-10: Web Track Ad-hoc Task

Term distillation for cross-db retrieval

In cross-DB retrieval, the domain of queries differs from the retrieval target in the distributio... more In cross-DB retrieval, the domain of queries differs from the retrieval target in the distribution of that of term occurrences. This causes incorrect term weighting in the retrieval system which assigns to each term a retrieval weight based on the distribution of term occurrences. To resolve the problem, we propose \term distillation" which is a framework for query term selection in cross-DB retrieval. The experiments using the NTCIR patent retrieval test collection demonstrate that term distillation is e ective for cross-DB retrieval.

Selecting expansion terms in automatic query expansion

Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval, 2001

University of Tokyo/Ricoh at Ntcir-3 Web Retrieval Task

In Proceedings of the …, 2002

In NTCIR-3 Web Task, we introduced new approaches in (1) similarity retrieval using one known rel... more

Term distillation in patent retrieval

Proceedings of the ACL-2003 workshop on Patent corpus processing -, 2003

In cross-database retrieval, the domain of queries di ers from that of the retrieval target in th... more In cross-database retrieval, the domain of queries di ers from that of the retrieval target in the distribution of term occurrences. This causes incorrect term weighting in the retrieval system which assigns to each term a retrieval weight based on the distribution of term occurrences. To resolve the problem, we propose \term distillation", a framework for query term selection in cross-database retrieval. The experiments using the NTCIR-3 patent retrieval test collection demonstrate that term distillation is e ective for cross-database retrieval.

NTCIR Advisor Report

NTCIR is one of the rst projects to construct, using a TREC-like contest operation, a practical J... more NTCIR is one of the rst projects to construct, using a TREC-like contest operation, a practical Japanese IR test collection. This report discusses design and the administration issues, such as the size and genre of the collection, the description of topics, tasks, the pre-test, and the collaboration with the IREX project.

A new character-based indexing method using frequency data for Japanese documents

Proceedings of the 18th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '95, 1995

... using Frequency Data for Japanese Documents OGAWA Yasushi and IWASAKI Masajirou {ogawa)iwasak... more

Transforming growth factor .beta.1: NMR signal assignments of the recombinant protein expressed and isotopically enriched using Chinese hamster ovary cells

Biochemistry, 1993

The transforming growth factor 8s are a homologous family of multifunctional cytokines that regul... more The transforming growth factor 8s are a homologous family of multifunctional cytokines that regulate cell growth and differentiation. As a prelude to studies of the solution structure and dynamics of TGF-81, we report virtually complete assignment of 'H and 15N resonances for this 25-kDa homodimeric protein. Recombinant TGF-81 was expressed in Chinese hamster ovary cells. The cells were grown either in a completely 15N-enriched medium or in a medium containing selectively 13C,15N-labeled amino acids to obtain either uniformly or specifically labeled protein, respectively. Two-and three-dimensional heteronuclear edited magnetic resonance spectra of the uniformly 15N-labeled protein and three samples selectively labeled with 13C and 15N yielded assignments for 96% of the backbone amide and C a protons and 87% of the side chain protons. To our knowledge, this is the first report of the use of an animal cell expression system to obtain extensive isotopic enrichment in order to sequentially assign a protein. The methodology described herein for the isotopic enrichment and resonance assignments of TGF-81 should be generally applicable to other eukaryotic proteins expressed by animal cells.

Ricoh at CLEF 2003

Lecture Notes in Computer Science, 2004

This paper describes the participation of RICOH in the Monolingual Information Retrieval tasks of... more This paper describes the participation of RICOH in the Monolingual Information Retrieval tasks of the Cross-Language Evaluation Forum (CLEF) 2003. We used our system with same kind of stemmer, same options and different parameters for 5 European languages to compare each result. Total performance of the system was reasonable. For French, German and Italian, we found some problems.

Lessons from BMIR-J2: a test collection for Japanese IR systems

... Haruo Kimoto (NTT) Ikuo Keshi (SHARP) Jun Toyoura (Mitsubishi Electric) Toshikazu Pukushima (... more

BMIR-J2: A Test Collection for Evaluation of Japanese Information Retrieval Systems

Sigir Forum, 1999

Term distillation for cross-db retrieval

The use of phrases from query texts in information retrieval

1. Introduction In this paper we examine the utility of phrases as search terms in information re... more

Supervised Synonym Acquisition Using Distributional Features and Syntactic Patterns

Journal of Natural Language Processing, 2009

Distributional similarity has been widely used to capture the semantic relatedness of words in ma... more Distributional similarity has been widely used to capture the semantic relatedness of words in many NLP tasks. However, parameters such as similarity measures must be manually tuned to make distributional similarity work effectively. To address this problem, we propose a novel approach to synonym identification based on supervised learning and distributional features, which correspond to the commonality of individual context types shared by word pairs. This approach also enables the integration with pattern-based features. In our experiment, we have built and compared eight synonym classifiers, and showed a drastic performance increase of over 60% on F-1 measure, compared to the conventional similarity-based classification. Distributional features that we have proposed are better in classifying synonyms than the conventional common features, while the pattern-based features have appeared almost redundant.

Optimizing query evaluation in n-gram indexing

Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval

In n-gram indexing widely used in Asian languages, because a long query word is divided into more... more In n-gram indexing widely used in Asian languages, because a long query word is divided into more than one n-gram, proximity constraints among ngrams need to be checked. And this checking degrades the retrieval speed. This paper proposes two optimizations of query evaluation: one selects a combination of n-grams whose processing cost is the minimum among all the possible combinations, and the other uses additional n-grams in finding potential documents whose proximity constraints need to be checked.

Simple word strings as compound keywords

Proceedings of the 16th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '93, 1993

This paper describes a new indexing method for Japanese text databases using the simple keyword s... more This paper describes a new indexing method for Japanese text databases using the simple keyword string, in which a compound word is treated as a string of simple words, which are the smallest units in Japanese grammar which still maintain their meanings. This method allows retrieved texts to be ranked, according to the similarity of their meaning to the query, without using a control vocabulary or thesaurus. This paper also introduces the keyword feature, which describes the syntactic and semantic characteristics of a word, and results in more precise keyword extraction and text retrieval as well as simple dictionary maintenance.

A New Indexing and Text Ranking Method for Japanese Text Databases Using Simple-Word Compounds as Keywords

Database Systems for Advanced Applications '93, 1993

This paper describes a new indexing method for Japanese text databases using the simpie keyword s... more This paper describes a new indexing method for Japanese text databases using the simpie keyword string. A compound word is treated as a string of simple words, which are the smallest units in Japanese grammar which still maintain their meanings. As a result, retrieved texts can be ranked, according to the similarity of their meaning and the query, without using a control vocabulary or thesaurus. For automatic indexing, the newly introduced keywordfeature, which describes the syntactic and semantic characteristics of a word, enables a precise keyword assignment and simplifies dictionary maintenance. Lexical and Syntactic Parsing: Since Japanese text contains no obvious space between (simple) words, the text must be lexically and syntactically parsed to identify simple words. Our system utilizes the Japanese parser developed at our Research CenterD]. A part of speech and keyword features, which are explained later, are also assigned to each simple word. Keyword Candidate Extraction: Based on the part of speech, simple words are extracted from the parsed text as keyword candidates.

Bucket Spreading Parallel Hash: a New, Robust, Parallel Hash Join Method for Data Skew In the Super Database Computer (SDC)

… of the sixteenth international conference on …, 1990

The Super Database Computer (SDC) is a high-performance relational database server for a join-int... more

RICOH at TREC-10: Web Track Ad-hoc Task

Term distillation for cross-db retrieval

In cross-DB retrieval, the domain of queries differs from the retrieval target in the distributio... more In cross-DB retrieval, the domain of queries differs from the retrieval target in the distribution of that of term occurrences. This causes incorrect term weighting in the retrieval system which assigns to each term a retrieval weight based on the distribution of term occurrences. To resolve the problem, we propose \term distillation" which is a framework for query term selection in cross-DB retrieval. The experiments using the NTCIR patent retrieval test collection demonstrate that term distillation is e ective for cross-DB retrieval.

Selecting expansion terms in automatic query expansion

Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval, 2001

University of Tokyo/Ricoh at Ntcir-3 Web Retrieval Task

In Proceedings of the …, 2002

In NTCIR-3 Web Task, we introduced new approaches in (1) similarity retrieval using one known rel... more

Term distillation in patent retrieval

Proceedings of the ACL-2003 workshop on Patent corpus processing -, 2003

In cross-database retrieval, the domain of queries di ers from that of the retrieval target in th... more In cross-database retrieval, the domain of queries di ers from that of the retrieval target in the distribution of term occurrences. This causes incorrect term weighting in the retrieval system which assigns to each term a retrieval weight based on the distribution of term occurrences. To resolve the problem, we propose \term distillation", a framework for query term selection in cross-database retrieval. The experiments using the NTCIR-3 patent retrieval test collection demonstrate that term distillation is e ective for cross-database retrieval.

NTCIR Advisor Report

NTCIR is one of the rst projects to construct, using a TREC-like contest operation, a practical J... more NTCIR is one of the rst projects to construct, using a TREC-like contest operation, a practical Japanese IR test collection. This report discusses design and the administration issues, such as the size and genre of the collection, the description of topics, tasks, the pre-test, and the collaboration with the IREX project.

A new character-based indexing method using frequency data for Japanese documents

Proceedings of the 18th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '95, 1995

... using Frequency Data for Japanese Documents OGAWA Yasushi and IWASAKI Masajirou {ogawa)iwasak... more

Transforming growth factor .beta.1: NMR signal assignments of the recombinant protein expressed and isotopically enriched using Chinese hamster ovary cells

Biochemistry, 1993

The transforming growth factor 8s are a homologous family of multifunctional cytokines that regul... more The transforming growth factor 8s are a homologous family of multifunctional cytokines that regulate cell growth and differentiation. As a prelude to studies of the solution structure and dynamics of TGF-81, we report virtually complete assignment of 'H and 15N resonances for this 25-kDa homodimeric protein. Recombinant TGF-81 was expressed in Chinese hamster ovary cells. The cells were grown either in a completely 15N-enriched medium or in a medium containing selectively 13C,15N-labeled amino acids to obtain either uniformly or specifically labeled protein, respectively. Two-and three-dimensional heteronuclear edited magnetic resonance spectra of the uniformly 15N-labeled protein and three samples selectively labeled with 13C and 15N yielded assignments for 96% of the backbone amide and C a protons and 87% of the side chain protons. To our knowledge, this is the first report of the use of an animal cell expression system to obtain extensive isotopic enrichment in order to sequentially assign a protein. The methodology described herein for the isotopic enrichment and resonance assignments of TGF-81 should be generally applicable to other eukaryotic proteins expressed by animal cells.

Ricoh at CLEF 2003

Lecture Notes in Computer Science, 2004

This paper describes the participation of RICOH in the Monolingual Information Retrieval tasks of... more This paper describes the participation of RICOH in the Monolingual Information Retrieval tasks of the Cross-Language Evaluation Forum (CLEF) 2003. We used our system with same kind of stemmer, same options and different parameters for 5 European languages to compare each result. Total performance of the system was reasonable. For French, German and Italian, we found some problems.

Lessons from BMIR-J2: a test collection for Japanese IR systems

... Haruo Kimoto (NTT) Ikuo Keshi (SHARP) Jun Toyoura (Mitsubishi Electric) Toshikazu Pukushima (... more

BMIR-J2: A Test Collection for Evaluation of Japanese Information Retrieval Systems

Sigir Forum, 1999

Term distillation for cross-db retrieval

The use of phrases from query texts in information retrieval

1. Introduction In this paper we examine the utility of phrases as search terms in information re... more