Filters








12 Hits in 3.0 sec

Lessons from BMIR-J2

Tsuyoshi Kitani, Tetsuya Sakai, Takenobu Tokunaga, Hiroshi Tsuruoka, Hidekazu Nakawatase, Teru Agata, Yasushi Ogawa, Tetsuya Ishikawa, Haruo Kimoto, Ikuo Keshi, Jun Toyoura, Toshikazu Fukushima (+2 others)
1998 Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '98  
BMIR-JP is the lirat complete Japanese test collection available for use in evaluating information retrieval systems.  ...  This collection and the principles derived in designing it should be helpful in the future development of new test collections.  ...  We thank Professor Katsumi Tanaka for his continuous support on this project. Thanks are also due to Noriko Kando who joined the discussion for the development.  ... 
doi:10.1145/290941.291033 dblp:conf/sigir/KitaniOIKKTFMUSTTNA98 fatcat:n6lyhfar6vbtxhlqxehrdvw6zm

Experiments in Japanese text retrieval and routing using the NEAT system

Gareth J. F. Jones, Tetsuya Sakai, Masahiro Kajiura, Kazuo Sumita
1998 Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '98  
Results on the standard BMIR-Jl and BMIR-J2 Japanese retrieval collections indicate that term weighting transfers well to Japanese text.  ...  This paper describes a structured investigation into the retrieval of Japanese text.  ...  . 4 Test Collections Ideally we would like to evaluate the NEAT system on large generally available Japanese test collections.  ... 
doi:10.1145/290941.290992 dblp:conf/sigir/JonesSKS98 fatcat:rgvygdng6bfidlirmagy6sllxe

A new approach to unsupervised text summarization

Tadashi Nomoto, Yuji Matsumoto
2001 Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '01  
To find the effectiveness of our approach under the proposed evaluation scheme, we set out to examine how a system with the diversity functionality performs against one without, using the BMIR-J2 corpus  ...  , a test data developed by a Japanese research consortium.  ...  TEST DATA AND EVALUATION PROCE-DURE BMIR-J2 BMIR-J2 (Benchmark for Japanese IR system version 2) developed jointly by a Japanese academic society and a government-funded research consortium represents  ... 
doi:10.1145/383952.383956 dblp:conf/sigir/NomotoM01 fatcat:uou7z2ox5rdbjn4kyuhy6tvhgu

The diversity-based approach to open-domain text summarization

Tadashi Nomoto, Yuji Matsumoto
2003 Information Processing & Management  
as BMIR-J2.  ...  To find the effectiveness of our approach under the proposed evaluation scheme, we set out to examine how a system with the diversity functionality performs against one without, using the test data known  ...  Test data and evaluation procedure BMIR-J2 BMIR-J2 (Benchmark for Japanese IR system version 2) represents a test collection of 5080 news articles in Japanese, all of them published in 1994 (Nichi-Gai  ... 
doi:10.1016/s0306-4573(02)00096-1 fatcat:oyxpckkxubhu7c25qyyd6igaoq

Spotting Where to Read on Pages - Retrieval of Relevant Parts from Page Images [chapter]

Koichi Kise, Masaaki Tsujino, Keinosuke Matsumoto
2002 Lecture Notes in Computer Science  
Based on the experimental results of retrieving Japanese newspaper articles, we have shown that the proposed method is superior to a method without the function of dealing with parts, and sometimes equivalent  ...  This paper presents a new method of document image retrieval that is capable of spotting parts of page images relevant to a user's query.  ...  We used BMIR-J2 based on the Mainichi Shinbun CD-ROM'94 data collection, as well as a Japanese morphological analyzer JUMAN.  ... 
doi:10.1007/3-540-45869-7_43 fatcat:kpakccjsu5dcrms5s7c4cg2gz4

CURE: Collection for Urdu Information Retrieval Evaluation and Ranking [article]

Muntaha Iqbal, Kamran Amjad, Bilal Tahir, Muhammad Amir Mehmood
2020 arXiv   pre-print
To the best of our knowledge, this work is the first attempt for preparing a standardized information retrieval evaluation test collection for the Urdu language.  ...  In this research work, we propose and construct a standard test collection of Urdu documents for IR evaluation and named it Collection for Urdu Retrieval Evaluation (CURE).  ...  Precision and recall measures were used for evaluation and reported average precision and recall of 52% and 82%, respectively. A benchmark for Japanese IR was developed and named as BMIR-J2 [25] .  ... 
arXiv:2011.00565v1 fatcat:5dmnuiz7wraafo65b2gu2sffsa

Cross-Language Chinese Text Retrieval in NTCIR Workshop

Kuang-hua Chen, Hsin-Hsi Chen
2001 SIGIR Forum  
The definitions, schedules, test collection (CIRB010), search results, evaluation, and initial analyses of search results of CHIR and ECIR are discussed in this article.  ...  The new plan of NTCIR towards multilingual Cross-Language Information Retrieval (CLIR) is also described.  ...  We are grateful to all pioneers in the area of IR evaluation for their efforts in paving a smooth way for followers.  ... 
doi:10.1145/511144.511149 fatcat:sg46hp56trfm7likgmsr2thpgi

Paraphrasing Japanese noun phrases using character-based indexing

Tokunaga Takenobu, Tanaka Hozumi, Kimura Kenji
2003 Proceedings of the second international workshop on Paraphrasing -  
This paper proposes a novel method to extract paraphrases of Japanese noun phrases from a set of documents.  ...  The proposed method consists of three steps: (1) retrieving passages using character-based index terms given a noun phrase as an input query, (2) filtering the retrieved passages with syntactic and semantic  ...  Experiments Data and preprocessing As input noun phrases, we used 53 queries excerpted from Japanese IR test collection BMIR-J2 1 (Kitani et al., 1998) based on the following criteria. • A query has  ... 
doi:10.3115/1118984.1118995 dblp:conf/acl-iwp/TokunagaTK03 fatcat:n26qcflr4reelhwizadn2izcmi

Graded Relevance [chapter]

Tetsuya Sakai
2020 Evaluating Information Retrieval and Access Tasks  
NTCIR was the first large-scale IR evaluation conference series to construct test collections with graded relevance assessments: the NTCIR-1 test collections from 1998 already featured relevant and partially  ...  In this chapter, I provide a survey on the use of graded relevance assessments and of graded relevance measures in the past NTCIR tasks which primarily tackled ranked retrieval.  ...  In particular, I thank the following past task organisers for clarifications regarding their overview papers: Cheng Luo, Yiqun Liu, and Takehiro Yamamoto (IMine), Makoto P.  ... 
doi:10.1007/978-981-15-5554-1_1 fatcat:zmifqjkcezgpximffpjxjpnklm

Graded Relevance Assessments and Graded Relevance Measures of NTCIR: A Survey of the First Twenty Years [article]

Tetsuya Sakai
2019 arXiv   pre-print
NTCIR was the first large-scale IR evaluation conference to construct test collections with graded relevance assessments: the NTCIR-1 test collections from 1998 already featured relevant and partially  ...  I then provide a survey on the use of graded relevance assessments and of graded relevance measures in the past NTCIR tasks which primarily tackled ranked retrieval.  ...  IR test collections BMIR-J1 and BMIR-J2 (which also had 3-point relevance levels) (Sakai et al, 1999) .  ... 
arXiv:1903.11272v1 fatcat:6lkqtg43svglnpeg7pu2fogyxe

BRIDJE over a language barrier

Tetsuya Sakai, Makoto Koyama, Masaru Suzuki, Akira Kumano, Toshihiko Manabe
2003 Proceedings of the sixth international workshop on Information retrieval with Asian languages -   unpublished
This paper describes two new features of the BRIDJE system for cross-language information access.  ...  Its advantage over a "black-box" machine translation approach is consistent across five test collections and across two language permutations: English-Japanese and Japanese-English.  ...  As with monolingual IR, CLIR evaluations usually rely on the use of static test collections: The system accepts a source language search request and outputs a ranked list of target language documents,  ... 
doi:10.3115/1118935.1118944 fatcat:blihiywvlnccxpa5iha25jnmcq

Texture Feature-based Document Image Retrieval

Fahimeh Alaei, University, My, Michael Blumenstein
2019
However, designing automation systems to accurately retrieve document images from archives remains a challenging problem.  ...  Finding discriminative and effective features is the fundamental task for developing an efficient retrieval system.  ...  This finding has important implications for developing a more efficient system for the collection of large size images.  ... 
doi:10.25904/1912/658 fatcat:wfs7xjfcvjcwvmiatmhg2ewngu