In this research, we conducted the experiment of automatically classifying the reference records in the Collaborative Reference Database of the National Diet Library into Nippon Decimal Classification (NDC). Three machine learning methods: decision tree, Naïve Bayes, and Support Vector Machine (SVM) methods - were tested on the words included in the sentences of questions and answers in 634 reference records. The results indicate that the using the SMV method and answer sentences were classified successfully compared to others. As for the recall and precision ratio, there was small difference among the methods, and between question and answer sentences. The significant difference, however, was observed in the recall ratio by the NDC category; the result of some category obtained more than 60% recall ratio, which suggested that classifying reference records into NDC by using machine learning methods was effective.

View full abstract

Co-citation, which is a typical similarity indicator among papers, has a premise that the degrees of similarity among cited papers are equal in one citing paper. But the similarity strength can be guessed from the distance between the places where cited papers are shown in a citing paper. In this paper, three distance measures ("physical distance", "co-occurrence of citing words", and "structural distance") are introduced for the purpose and these measures are evaluated.

View full abstract

This paper reports an information retrieval system for Japanese bibliographic data based on SRU/SRW protocol. The system consists of a client which can access to our SRU/SRW server, and of a server which provides bibliographic data for educational papers at Library of Education. The server has about 120 thousands bibliographic records of "Kyoiku-Kenkyu-Ronbun-Sakuin", and are capable for basic retrieval functions in SRU/SRW, such as Boolean queries, queries for specific search indexes, etc. Its capabilities are confirmed by using two kinds of query datasets, functional test-set and OPAC log-based one, in a batch search experiment.

View full abstract

A researcher studies a new target domain through a search of papers. However, an efficient search is difficult if the information demands are ambiguous. To solve this problem, it is necessary prior to a search of the papers to have an understanding of the entirety of the research field and an overview of the research topics. In this paper, we propose a technique to extract the research topics by using author-keyword relationship in a specified research area. And we introduce the prototype system based on our proposal technique.

View full abstract

To act for the solution of the global problems, as the damages of environment and ecosystem of the earth, the shortage of energy and resources, we should not depend on using skill of trade-off which enables compromise conflicting values of different behaviors of life, but we should make up the structure of Lattice which means fusion of different values of different behaviors of life. And to attain the purpose, we have to evolve our abilities of realizing Abduction.

View full abstract

This article dealt with a music copyright. Mass consumption of the musical works gave big influence to the copyright system, and a system to control a copyright intensively was made. I chase a process of the circulation of the musical works, the use and generation and the development of the music copyright from the legal system and the social side. This article passes through the Revolution period from Ancien régime in France, and to reach the establishment of SACEM in the 19th century. Therefore it is considered the history of the score publication, a French copyright system, the sudden rise of the management system of the music copyright.

View full abstract

At Japan Science and Technology Agency (JST), we are developing a new function "Analysis and Visualization" for our document retrieval system "JDreamII". In this report, I describe the frequency analysis function on the current JDreamII, and Analysis and Visualization using our organization dictionary. Future developments are also described.

View full abstract

The information on the Web is increasing explosively, and the Internet/Web has become the information infrastructure of our own time. Web users gather and find information through search engines using keywords related to their interests. However, it is not always the case that users conceive suitable keywords relevant to the problem in question. In this paper, we propose a novel method for query expansion based on the user profile generated from browsing histories of a user. This enables a user the context-sensitive keyword expansion using profiles specific to the user. Here, we assume that a user's web browsing behavior corresponds to the user's intended query. We illustrate the usefulness of our proposed method for query expansion with some examples.

View full abstract

The metadata elements were extracted from the explanations of Japanese Classics on the web of Tsurumi University Library. They were divided into several groups and compared among four types of Japanese Classics, the manuscript, the printed book, the detached segment, and the other. It presents a clear difference of the metadata elements among four types of Japanese Classics. We discussed the difference of them from an aspect of domain knowledge.

View full abstract

The documents and scriptures in the possession of temples are a key to making clear the transcription of the writings, the human relationships and the religious circumstances behind them, by means of the colophons that state the background of themselves. We have been developing the applications for supporting the activities of humane studies. In this paper, we introduce a scripture retrieval system powered by the full-text search engine Hyper Estraier, and an improved version of genealogical visualization system which allows any priest to have more than one master in a genealogical table. In addition the prospect of the cooperation between these two applications is discussed.

View full abstract

The system of Grants-in-Aid for Scientific Research form Ministry of Education, Culture, Sports, Science and Technology of Japan is one of the oldest funding systems, which has been supporting researchers belonging to universities and research institutes in Japan. This paper examines the researchers' need of funding chronologically through their career stages by the amount of grants awarded.

View full abstract

The system of Grants-in-Aid for Scientific Research from MEXT is one of the oldest ones, which is the funding system for researchers belonging to universities and institutes in Japan. In this Grants-in-Aid, it is very interesting to see how the adoption situation of the research of these prioritized areas on the 2nd Science and Technology Basic Plan has changed. We measure strength of a related level each other by comparing "Field feature keyword" that has been developed so far and the research subjects and the applied detail research fields in the Grants-in-Aid. Especially in this analysis, the relation between a field related to the genome, a nanotechnology in the prioritized area and the applied research subjects and their applied detail research fields are examined by using this method, and the secular distortion of the adoption situation and the mutual relation are examined by using Correspondence Analysis etc.

View full abstract

One of the characteristics of digital archives used to digitize and present historical materials and cultural resources is that they allow you to handle materials of various types and content synthetically. In addition, putting together various materials into one archive allows the possibility to find new relationships between materials that have been conventionally separated by type, as well as enable you to search materials in a cross-sector manner. Notably materials created in the same era may include common information through geological condition, figures, social organizations and historical events.
In this study, materials are integrated through an understanding of the relationship between various historical artifacts with ontology. An integrated digital archive is constructed based on these viewpoints and verified.

View full abstract

In terminology theories, hierarchical relationship between concepts is regarded to be central. After reviewing the classical theories and the criticism against them, the discussions and the points on the hierarchical relation are described, focusing on the individual concept, the partitive relation, and the context dependency of the hierarchy. The followings are concluded. In terminology, the hierarchical relations are subsequent relations resulting from the construction of a concept system, and therefore make few serious confusion in the terminology workplace even though they are inappropriately defined, compared to thesauri for information retrieval. The disciplinary view is reflected on the selection of essential characteristics. The arbitrariness and the intuitiveness are hardly avoidable in hierarchical relations. The exploitation of the viewpoint of hierarchical relation may be of use during the course of exploring the 'concept of concept' in terminology science, which is indispensable for the wider and multicultural application of the terminology theory.

View full abstract