Faculty Profiles - OHTA Manabu

MISC - OHTA Manabu

Division display 81 - 100 of about 112 ／ All the affair displays >>

A Prediction Search Using Topic Distillation by Document Clustering

2010 ( 2 ) 1 - 8 2010.11

A Browsing Support System of Research Papers Using Web Resources

HACHIKI TOSHIHIRO, OHTA MANABU, TAKASU ATSUHIRO

2009 ( 14 ) 1 - 6 2009.11

D-12-48 Bibliographic Element Extraction from OCRed Text of Academic Journals

Yakushi Takayuki, Ohta Manabu, Takasu Atsuhiro

Proceedings of the IEICE General Conference 2009 ( 2 ) 157 - 157 2009.3

D-024 Dynamic Personalized Clustering for Web Search Results Based on Users' Familiarity with the Topics

Aono Hiroshi, Ohta Manabu

7 ( 2 ) 99 - 102 2008.8

D-044 Selection methods of relevant Web pages for blog clustering

Toda Yuji, Ohta Manabu

6 ( 2 ) 107 - 109 2007.8

On difference-conscious clustering and its feature aggregation methods

WATANABE Masashi, OHNO Shigeyoshi, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IPSJ SIG Notes 2005 ( 67 ) 199 - 206 2005.7

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

Clustering is very useful analyzing a huge amount of data. It is preferable that we can understand relationships between clusters intuitively in order to reach necessary information within the clustering results. Then, we propose a method of analyzing differences between clusters by using visualization. We a two-dimensional map display based on MDS(MultiDimensional Scaling) for visualization in this article. In addition, we examine two methods for aggregating the features used for clustering, one of which is based on the tf-idf weight of feature terms and the other on a rank aggregation method called the MEDRANK.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00019114/

Clustering Web Pages Based on Maximum Flow Algorithm

OHNO Shigeyoshi, WATANABE Masashi, KATAYAMA Kaoru, ISHIKAWA Hiroshi, OHTA Manabu

IEICE technical report. Data engineering 105 ( 171 ) 65 - 70 2005.7

Language：Japanese Publisher：The Institute of Electronics, Information and Communication Engineers

While search engines are indispensable for searching on the Web, users have to check a long ordered list to locate the needed information. It is often tedious and less efficient. In this paper, we propose a new link-based clustering approach to categorize search results returned from Web search engine. The maximum flow algorithm effective in finding the page sets with the link structures connected more precisely is used for the analysis of link information. In the experiments to evaluate method perfoemance quantitatively, we had good results using the data of NTCIR.

CiNii Article

CiNii Books

researchmap

On difference-conscious clustering and its feature aggregation methods

WATANABE Masashi, OHNO Shigeyoshi, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IEICE technical report. Data engineering 105 ( 171 ) 155 - 160 2005.7

Language：Japanese Publisher：The Institute of Electronics, Information and Communication Engineers

Clustering is very useful analysing a huge amount of data. It is preferable that we can understand relationships between clusters intuitively in order to reach necessary information within the clustering results. Then, we propose a method of analyzing differences between clusters by using visualization. We a two-dimensional map display based on MDS(MultiDimensional Scaling) for visualization in this article. In addition, we examine two methods for aggregating the features used for clustering, one of which is based on the tf-idf weight of feature terms and the other on a rank aggregation method called the MEDRANK.

CiNii Article

CiNii Books

researchmap

A Proposal of a Static Load Distribution Technique on P2P Network

SHIOYA Yasuo, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IPSJ SIG Notes 2004 ( 72 ) 675 - 682 2004.7

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

In distributed computing, when a load is distributed uniformly , the processing time varies with node, and the total processing time will be dependent on the maximum processing time on each node. Therefore distributing load according to the capability of each node can make us expect more efficient total processing by equalization of load processing time on each node. This paper defines a "rank" as a metric which indicates the throughput of each node, and shows how to determine the "rank" of each node from "ranks" of adjoining node. Moreover, this paper proposes a technique to distribute load based on a "rank" of a node.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00019209/

A Page Fusion Method using Case Grammar

WATANABE Takuya, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IPSJ SIG Notes 2004 ( 72 ) 653 - 660 2004.7

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

Multi-document summarization techniques are researched to get information efficiently from a huge amount of text data. We have already proposed techniques which use morphological analysis and dependency structure analysis, extract semantic information from documents and integrate it into pseudo natural language. In this article, combined this Multi-Document Fusion method and case grammar, and aimed at an improvement of fusion precision and readability. If understanding a word concept is required to analyze deep case, we guess it using search engine's hit counts.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00019206/

SAX - GTR : Fast Loading of XML Stream

YOKOYAMA Shohei, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IPSJ SIG Notes 2004 ( 71 ) 213 - 220 2004.7

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

There are two basic types of XML parser: SAX and DOM. The SAX is an event driven XML parser that parses an XML input stream. The DOM is an in-memory tree representation of the XML document. Unlike SAX, DOM is a read-write API. Documents can be searched, queried, and updated via the DOM interface. This makes DOM much more effective when randomly accessing to XML documents. However, it is quite memory exhaustive compared with SAX because SAX doesn't have to store the entire document in memory. Therefore SAX is scalable from small to very large XML documents. The time taken to parse an XML document grows only linearly with the size of the document. In this paper, we propose an efficient framework for SAX applications called SAX-GTR that divides an XML document in two binary files; one is for the event stream and frequently-accessed data, the other is for non-frequently-accessed data. The division is decided automatically with a particular SAX event handler, without the structural and semantic information such as DTDs. The purpose of this division is to decrease the cost of loading. We also present a transparent access to SAX-GTR documents using SAX event handlers. The effectiveness of SAX-GTR is demonstrated by some experiments.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00019243/

A Page Fusion Method using Case Grammar

WATANABE Takuya, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IEICE technical report. Data engineering 104 ( 178 ) 49 - 54 2004.7

Language：Japanese Publisher：The Institute of Electronics, Information and Communication Engineers

CiNii Article

CiNii Books

researchmap

A Proposal of a Static Load Distribution Technique on P2P Network

SHIOYA Yasuo, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IEICE technical report. Data engineering 104 ( 178 ) 67 - 72 2004.7

Language：Japanese Publisher：The Institute of Electronics, Information and Communication Engineers

CiNii Article

CiNii Books

researchmap

SAX-GTR : Fast Loading of XML Stream

YOKOYAMA Shohei, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IEICE technical report. Data engineering 104 ( 176 ) 169 - 174 2004.7

Language：Japanese Publisher：The Institute of Electronics, Information and Communication Engineers

CiNii Article

CiNii Books

researchmap

An Interactive Annotation Framework for Image Retrieval

LOISANT Erwan, MARTINEZ Jose, ISHIKAWA Hiroshi, OHTA Manabu, KATAYAMA Kaoru

IEICE technical report. Data engineering 104 ( 176 ) 211 - 216 2004.7

Language：English Publisher：The Institute of Electronics, Information and Communication Engineers

Image retrieval is usually based on content information, external annotation, or a combination of the two. While content information is intrinsic to the image and thus is an objective data, external annotation may imply human and be deeply dependent on the culture and sensibility of the annotator. Consequently, two operators would not produce the same annotation for the same image, and both results may not be satisfying for user. However, an image usually contain important information impossible to retrieve from content, thus relying only on content information does not produce satisfying results. To answer to this issue, we propose an annotation framework featuring an explicit separation of objective and subjective informations, objective informations including content based information and a limited number of annotations, the other annotations being considered as subjective information and collected with the help of the user. We also propose an efficient way to apply these subjective annotations to a navigation structure calculated before-hand from the objective informations. The resulting browsing technique brings the relevance of human annotation without the cost of a before-hand annotation, and without denying the robustness of a before-hand calculated structure.

CiNii Article

CiNii Books

researchmap

Comparison Between Multi - Vector Feature Space Method and Earth Mover's Distance Method In Similarity Searches of Images

Yamane Yasuo, Hoshiai Tadashi, Tsuda Hiroshi, Katayama Kaoru, Ohta Manabu, Ishikawa Hiroshi

IPSJ SIG Notes 2004 ( 45 ) 83 - 90 2004.5

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

There are some cases where dissimilar images are judged to be similar in using quadratic-form distance, a representative one in similarity searches of images. To address this problem, we proposed a multi-vector feature space method based on pseudo-Euclidean space and an oblique basis (MVPO). In this method, an image is represented by a solid consisting of multiple vectors; each vector corresponds to each feature. We also proposed D-distance as a distance between solids. A representative method similar to ours is Earth Mover's Distance (EMD), which is said to experimentally outperforms others including quadratic-form distance in precision. We show that EMD can be formalized as a distance in MVPO, and a reason why it outperforms quadratic-form distance. We also mention the difference between D-distance and EMD.

CiNii Article

CiNii Books

researchmap

Overlapping Clustering Method Using Local and Global Importance of Feature Terms at NTCIR-4 WEB Task

Manabu Ohta, Hirokazu Narita, Shigeyoshi Ohno

Proceedings of the Fourth NTCIR Workshop on Research in Information Access Technologies Information Retrieval, Question Answering and Summarization(NTCIR) 2004

On the Evaluation of Clustering-Based Web Search Engines

Mizusawa Naoto, Ohta Manabu, Katayama Kaoru, Ishikawa Hiroshi

Proceedings of the IEICE General Conference 2003 ( 1 ) 46 - 46 2003.3

A proposal of modified P2P network using query history. ~Application to the safety information retrieval.~

2002 ( 67 ) 327 - 334 2002.7

Meta Search Engine Using Hierarchical Clustering : METAL

NARITA Hirokazu, OHTA Manabu, KATAYAMA Kaoru, ISHIKAWA Hiroshi

IPSJ SIG Notes 2002 ( 67 ) 375 - 382 2002.7

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

It is search engines represented by google, goo, Lycos, and so on that shoulders a leading role as an access means to the necessary information on the Web. A lot of maintenance, such as collection of new Web pages and renewal check, is necessary for these search engines to cope with the Web changing continually. On the other hand, sometimes, it is difficult to find necessary information because of enormous search results. One search engine may be able to find what another can't because different search engines collect different kinds of information. In this paper, we propose a meta search engine which uses more than one search engine at the same time for improving recall, and cluster search results for supporting users to find their necessary information more easily.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00019469/

　 PREV - NEXT 　

< OHTA Manabu >

MISC - OHTA Manabu