88 Hits in 4.8 sec

English-to-Japanese Cross-Language Question-Answering System using Weighted Adding with Multiple Answers (pp. 17-22)

Masaki Murata, Masao Utiyama, Toshiyuki Kanamaru, Hitoshi Isahara
2009 POLIBITS Research Journal on Computer Science and Computer Engineering With Applications  
We describe a method of using multiple documents with decreasing weights as evidence to improve the performance of a question-answering system.  ...  Because simply adding scores degrades the performance of question-answering systems, we add scores with decreasing weights to reduce the negative effect of simply adding.  ...  USE OF MULTIPLE DOCUMENTS AS EVIDENCE WITH DECREASED ADDING Suppose that the question, "What is the capital of Japan?"  ... 
doi:10.17562/pb-40-3 fatcat:mivoqyqvjjhcxki7huwz45y3q4

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  ...  In this paper, I first describe a few graded-relevance measures that originated from NTCIR (and a few variants) which are used across different NTCIR tasks.  ...  CQA (NTCIR-8) The NTCIR-8 Community Question Answering (CQA) task (Sakai et al, 2010a ) was an answer ranking task: given a question from Yahoo! Chiebukuro (Japanese Yahoo!  ... 
arXiv:1903.11272v1 fatcat:6lkqtg43svglnpeg7pu2fogyxe

Evaluation of Information Access with Smartphones [chapter]

Makoto P. Kato
2020 Evaluating Information Retrieval and Access Tasks  
NTCIR 1CLICK and MobileClick are the earliest attempts toward testcollection-based evaluation for information access with smartphones.  ...  This chapter begins with an introduction to NTCIR 1CLICK and MobileClick, explains the evaluation methodology and metrics such as S-measure and M-measure, and finally discusses the potential impacts of  ...  Finally, I would like to express my special thanks to Stefano Mizzaro, who reviewed the initial version of this chapter and provided many useful suggestions.  ... 
doi:10.1007/978-981-15-5554-1_11 fatcat:ebeg3rb525aurezbjoydkqkapm

Leveraging social Q&A collections for improving complex question answering

Youzheng Wu, Chiori Hori, Hideki Kashioka, Hisashi Kawai
2015 Computer Speech and Language  
Answers as knowledge repositories and investigates techniques to mine knowledge from them to improve sentence-based complex question answering (QA) systems.  ...  We compare our approach with the question-specific and monolingual translation-based methods presented in previous works.  ...  Table 5 lists some of the questions used with their best answers as basic positive training pairs of the corresponding type of complex questions.  ... 
doi:10.1016/j.csl.2014.06.001 fatcat:3ljbbiiskrg4hmxpkzxecrywma

Modeling anchor text and classifying queries to enhance web document retrieval

Atsushi Fujii
2008 Proceeding of the 17th international conference on World Wide Web - WWW '08  
We use the NTCIR test collections and show the effectiveness of individual methods and the entire Web retrieval system experimentally.  ...  Our retrieval system combines the results obtained with the content-based and anchor-based retrieval methods, in which the weight for each retrieval result is determined automatically depending on the  ...  The differences of ADM and AAM for NTCIR-4 and NTCIR-5 were significant at the 5% and 1% levels, respectively.  ... 
doi:10.1145/1367497.1367544 dblp:conf/www/Fujii08 fatcat:4meebunwabgx5eozfq3rekcvgq

Improving Keyword Recognition of Spoken Queries by Combining Multiple Speech Recognizer's Outputs for Speech-driven WEB Retrieval Task

2005 IEICE transactions on information and systems  
Comparing with the retrieval accuracies when an LM with a 20,000/60,000 vocabulary size is used in an LVCSR system, we found that the larger the vocabulary size is, the better the retrieval accuracy is  ...  Thus, we regard an LVCSR model as one of various LVCSR systems if decoders used among the LVCSR systems are the same, but the type of acoustic model is different from others.  ...  Katsunobu Itoh (Nagoya University) for providing the trigram and bigram language models with 60,000 vocabulary size.  ... 
doi:10.1093/ietisy/e88-d.3.472 fatcat:jrjmblechbgyjgdkayojwpl5cy

Associative document retrieval by query subtopic analysis and its application to invalidity patent search

Toru Takaki, Atsushi Fujii, Tetsuya Ishikawa
2004 Proceedings of the Thirteenth ACM conference on Information and knowledge management - CIKM '04  
By exploiting certain unique features of Japanese patent claims, we use features distinguishing the preamble and the essential portion in a query patent claim.  ...  For each subtopic element, a sub-query is produced and similar documents are retrieved with a relevance score.  ...  With the Boolean systems, no ranking among the retrieved documents is provided. In another type of patent IR system (e.g., NRI Cyber Patent 5 ), a natural language text is used as a query.  ... 
doi:10.1145/1031171.1031251 dblp:conf/cikm/TakakiFI04 fatcat:5ba52qn3jzeetoqze67phezqia

Automatically Acquiring Causal Expression Patterns from Relation-annotated Corpora to Improve Question Answering for why-Questions

Ryuichiro Higashinaka, Hideki Isozaki
2008 ACM Transactions on Asian Language Information Processing  
This paper describes our approach for answering why-questions that we initially introduced at NTCIR-6 QAC-4.  ...  NAZEQA, a Japanese why-QA system based on our approach, clearly outperforms baselines with a Mean Reciprocal Rank (top-5) of 0.223 when sentences are used as answers and with a MRR (top-5) of 0.326 when  ...  NAZEQA, a Japanese why-QA system based on our approach, clearly outperforms baselines with a MRR (top-5) of 0.223 when sentences are used as answers and with a MRR (top-5) of 0.326 when paragraphs are  ... 
doi:10.1145/1362782.1362785 fatcat:kfhgxxkikffqdoujxzgeicvmky

Expanding Paraphrase Lexicons by Exploiting Generalities

Atsushi Fujita, Pierre Isabelle
2018 ACM Transactions on Asian and Low-Resource Language Information Processing  
The results of our experiments on English, French, and Japanese demonstrate that our method manages to expand seed lexicons by a large multiple.  ...  Vous avez des questions? Nous pouvons vous aider.  ...  [48] have collected multiple human translations of the same source text. Multiple verbalizations of mathematical proofs have also been used [5] .  ... 
doi:10.1145/3160488 fatcat:tvpx2bw4tbdcxdilzs4jfwvr4m

Cross-language Information Retrieval [article]

Petra Galuščáková, Douglas W. Oard, Suraj Nair
2021 arXiv   pre-print
This chapter reviews the state of the art for cross-language information retrieval and outlines some open research questions.  ...  CLIR tasks oriented at Japanese and eastern Asian languages were first organized at what is now called the NII Testbeds and Community for Information access Research (NTCIR) evaluations between 2002 and  ...  to provide a direct answer to a question.  ... 
arXiv:2111.05988v1 fatcat:fgnaux4lcbe5jlpczhbxka5cqq

A clinical specific BERT developed with huge size of Japanese clinical narrative [article]

Yoshimasa Kawazoe, Daisaku Shibata, Emiko Shinohara, Eiji Aramaki, Kazuhiko Ohe
2020 medRxiv   pre-print
Approximately 120 millions of clinical text stored at the University of Tokyo Hospital were used as dataset.  ...  In this work, we demonstrate a development of a clinical specific BERT model with a huge size of Japanese clinical narrative and evaluated it on the NTCIR-13 MedWeb that has pseudo-Twitter messages about  ...  Methods Datasets Approximately 120 million lines of clinical text for eight years stored at electronic health record system of the University of Tokyo Hospital were used.  ... 
doi:10.1101/2020.07.07.20148585 fatcat:d7sur7q3gbgxbcr6d5oe74xkvm

Information access across languages on the web: From search engines to digital libraries

Jiangping Chen, Yu Bao
2009 Proceedings of the American Society for Information Science and Technology  
They adapted 50 topics that have been evaluated at NTCIR-5 Cross-Lingual Information Retrieval Task (  ...  CLIA research has been vigorously pursued through TREC (, CLEF (, an Asian Language Retrieval and Question-answering Workshop called NTCIR (  ... 
doi:10.1002/meet.2009.1450460278 fatcat:knndrpbqgjcrhicunmvjhwh7ii

Bridging the Language Gap: Learning Distributed Semantics for Cross-Lingual Sentiment Classification [chapter]

Guangyou Zhou, Tingting He, Jun Zhao
2014 Communications in Computer and Information Science  
Sentiment classifiers trained on the source language can be adapted to predict sentiment polarity of the target language with the language-independent distributed representations.  ...  In this paper, we propose a new model which uses stacked autoencoders to learn language-independent distributed representations for the source and target languages in an unsupervised fashion.  ...  MPQA-EN (Labeled English Data): The multi-perspective question answering (MPQA-EN) corpus [25] consists of newswire documents manually labeled with subjectivity information.  ... 
doi:10.1007/978-3-662-45924-9_13 fatcat:raj3ar3rkfcudojybu5fagt4ni

Report on the SIGIR 2007 workshop on focused retrieval

Andrew Trotman, Shlomo Geva, Jaap Kamps
2007 SIGIR Forum  
Question Answering has been examined by TREC, CLEF, and NTCIR for many years, and is arguably the ultimate goal of semantic web research for interrogative information needs.  ...  [7] examine question-type identification from multi-sentence queries in Japanese.  ... 
doi:10.1145/1328964.1328981 fatcat:3qi4wek47zffha4k7l2y5xkw4m

Patent Retrieval: A Literature Review [article]

Walid Shalaby, Wlodek Zadrozny
2017 arXiv   pre-print
With the ever increasing number of filed patent applications every year, the need for effective and efficient systems for managing such tremendous amounts of data becomes inevitably important.  ...  These tools must also be developed in hand with patent professionals considering their practices and expectations.  ...  The answers of these questions are not deterministic and probably require multiple interaction cycles with the user in order to satisfy his/her information need.  ... 
arXiv:1701.00324v1 fatcat:u5w55z4cj5cwbii5s55rjvegvi
« Previous Showing results 1 — 15 out of 88 results