Automatic summarization of Japanese sentences and its application to a WWW KWIC index

Kiyota, Yoji; Kurohashi, Sadao

doi:10.1109/saint.2001.905175

Cited by 8 publications

(7 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…After the convergence, the summary is generated using the following preference filtering strategy: 5 Step 1: Initialize the set of extracted sentencesS:: 1 and the set of keywords as K::={,@ Let Nmax be the upper bound for the number of sentences to be extracted.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A neural network-based automatic summarization for the minutes of local assemblies

Fujioka

Nishikawa

2006

2006 IEEE International Conference on Systems, Man and Cybernetics

View full text Add to dashboard Cite

An automatic summarization system for local assembly minutes is described. Our system is unique in the following two points: First, the completeness that puts emphasis on the readability of the generated summary is realized. Secondly, thepreference that prioritizes the selection based on the viewpoint on the content of the summary is considered. Our system has a t wo-level structure that brings about the property of "anytime algorithm": The neural network in the lower layer recollects important sentences based on the frequency ofthe words in the text. Then, based on the snapshot of the neural network, the summary gnerator in the upper layer extracts a cluster of sentences around the important sentences, with the prioritization by the user's preferred viewpoint. The effectiveness of our method is demonstrated using the actual minutes of a prefecture in Japan.

show abstract

Section: Methodsmentioning

confidence: 99%

“…The work [5] employs the extraction approach with the improvement by the use ofmorphological analysis and a parser and by making the unit for importance scoring more granular from a sentence to parts of a sentence. In [6], the transition of topic within a document is considered by clustering sentences based on the keywords.…”

Section: Introductionmentioning

confidence: 99%

A neural network-based automatic summarization for the minutes of local assemblies

Fujioka

Nishikawa

2006

2006 IEEE International Conference on Systems, Man and Cybernetics

View full text Add to dashboard Cite

show abstract

“…For example, de is known to have more than ten usages (e.g., by, for, with, at, etc.) [Ishiwata 1999;Kiyota and Kurohashi 2001], and this ambiguity makes it difficult to decide when it is used for a causal relation. The BACT patterns seem to use the semantic categories around functional words to distinguish contexts in which de can be used as a causal cue.…”

Section: Impact Of the Featuresmentioning

confidence: 99%

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

Higashinaka

Isozaki

2008

ACM Transactions on Asian Language Information Processing

View full text Add to dashboard Cite

This paper describes our approach for answering why-questions that we initially introduced at NTCIR-6 QAC-4. The approach automatically acquires causal expression patterns from relationannotated corpora by abstracting text spans annotated with a causal relation and by mining syntactic patterns that are useful for distinguishing sentences annotated with a causal relation from those annotated with other relations. We use these automatically acquired causal expression patterns to create features to represent answer candidates, and use these features together with other possible features related to causality to train an answer candidate ranker that maximizes the QA performance with regards to the corpus of why-questions and answers. 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 paragraphs are used as answers, making it presumably the best-performing fully implemented why-QA system. Experimental results also verified the usefulness of the automatically acquired causal expression patterns.

show abstract

“…In prior research on HTML text summarization [3,4], Web pages have been summarized by methods such as probabilistic models or combinations of syntax analysis and the TFIDF method [5]. However, they deal only with the sentences in the HTML text and ignore parts that have short word links or items listed, and thus they generate sentence-specific summaries.…”

Section: Introductionmentioning

confidence: 99%

HTML text segmentation for Web page summarization by a key sentence extraction method

Sunayama

Iyama

Yachida

2006

Systems & Computers in Japan

View full text Add to dashboard Cite

SUMMARYThe information displayed as the search result by search engines is important for quickly finding the desired information. In particular, the summary of each Web page in the search results is important for determining the Web page content, as well as for determining how the input search term is used in each Web page, namely, the relation between the search term and the Web page. However, the summaries of the search results in conventional search engines have problems such as extracting only the opening text and not containing the search term, or containing the search term but having the sentence truncated in the middle so that the context of the term or the content of the Web page cannot be determined. Therefore, a summary in sentence units is desirable, but since HTML text includes many nonsentence items that do not contain punctuation, if they are unprocessed, it is difficult for a key sentence extraction system that treats sentences as units to provide a summary. Thus, in this paper, we propose an HTML text segmentation system that divides the source text of each Web page into meaningfully connected groups of text corresponding to sentences. We also verify experimentally that the text generated by this system can be used effectively in a Web page summarization.

show abstract

Automatic summarization of Japanese sentences and its application to a WWW KWIC index

Cited by 8 publications

References 6 publications

A neural network-based automatic summarization for the minutes of local assemblies

A neural network-based automatic summarization for the minutes of local assemblies

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

HTML text segmentation for Web page summarization by a key sentence extraction method

Contact Info

Product

Resources

About