The Journal of Information Science and Technology Association
(Johono Kagaku to Gijutsu)
Abstract

Akira IDE, Ph.D
Kinki Univ. School of Economics
3-4-1 Kowakae, Higashi-Osaka City 577-8502 JAPAN

Abstract：In this paper, filtering system which tends to be taken as a technical term is analyzed in terms of social system. The purpose of filtering is to foster sound development of the youth, and this objective is reasonable. However, if we introduce filtering system without consideration, this would mean bringing new censorship into the world. Regarding the Internet as the one most important infrastructure for democracy, critical analysis of filtering system will be introduced in this paper for fear that it may destroy the concept of "freedom of expression" which is the base of democracy. Finally, I will suggest new policies for filtering system in the future.

Keywords：  filtering / democracy / freedom of expression / censorship / search engine

Toshihiro KAMISHIMA
National Institute of Advanced Industrial Science and Technology (AIST)
AIST Tsukuba Central 2
Umezono 1-1-1, Tsukuba, Ibaraki 305-8568 Japan

Abstracts：We are recently encountering the situation where we cannot find out our target information even if we have an access to it, because it is buried in so enormous amount of irrelevant information. This situation is known as information overload. Recommender systems are developed to solve this problem of information overload. These systems help for users to find out their target information. In this article, we overview the algorithms for making recommendations.

Keywords：  information overload / recommender system / content-based filtering / collaborative filtering / serendipity

Tetsuya ISHIKAWA^*1，Takayuki UDA^*2
*1Tetsuya ISHIKAWA
The Historiographical Institute, University of Tokyo
7-3-1 Hongo, Bunkyo-ku Tokyo, 113-0033 JAPAN
^*2Takayuki UDA
Graduate School of Information Sciences, TOHOKU University
2-1-1 Katahira, Aoba-ku, Sendai, 980-0812 JAPAN

Abstract：Any kind of information providing service systems are operating with the Internet widely using. In view of the situation, the type of information providing service system is able to device the two kind of type systems - Push-typed system and Pull-typed system. In the rapid spread of the electric commerce, as one of the information providing service system, which is the recommender system for providing any kind of goods/events' information. The basic system function of the recommender system is constructed on the filtering function. Then in this paper, the information filtering function is clearly, and the information recommendation system based on the collaborative filtering function is introduced.

Keywords：  information service system / information filtering / recommendation system / collaborative filtering / hybrid filtering

Toshihiro TABATA
Graduate School of Natural Science and Technology, Okayama University
3-1-1 Tsushima-naka, Okayama 700-8530 JAPAN

Abstract：SPAM mail has being increased with the spread of e-mail. The increase of SPAM mails becomes a serious problem for email users. Email filtering that can filter out SPAM mail, has been received much attention. Bayesian filter is one application of Bayes' Theorem. The filter program creates a database from received emails and calculates the SPAM probability of an email based on the database. This article explains the method of learning and calculation of SPAM probability of Bayesian filtering.

Keywords：  Bayesian filter / e-mail / SPAM mail / filtering / learning

Emi ISHIDA
Surugadai University, Faculty of Cultural Information Resources
698 Azu, Hanno-shi, Saitama 357-8555 JAPAN

Abstract：Text classification has two variations. One is text clustering, which is a way to group text according to its character. The other is text categorization. This task is to classify text to the predefined category. Text categorization technique is similar to the filtering. 　In this paper, I focus on text categorization. The processes of text categorization such as feature extraction, feature selection and building classifiers are briefly explained. I introduce test collections which are used for text categorization system evaluation. Finally, I suggest a new research topic in this area.

Keywords：  text classification / text categorization / clustering / classifiers / information retrieval / machine learning / test collection

Fumio SHIMPO
University of Tsukuba
1-2 Kasuga, Tsukuba, Ibaraki-Pref 305-8550 JAPAN

Abstract：Discussions concerning the use of filtering tends to examine a measure relating to the filtering of unlawful and/or harmful information. In general, when filters do not function properly, blame is placed on the technical side. This paper focuses on a survey of filtering from a legal perspective. Filters are used not only to block information but also to restrict access to the Internet and protect confidential information.

Keywords：  filtering / privacy / data protection / secrecy of communication / freedom of expression / information security

Masaaki KASHIMURA
HUMI Project, Keio University
2-15-45 Mita, Minato-ku, Tokyo 108-8345 JAPAN

Abstract：At a photographing session of a bound book for digitization, the main work is to photograph the frontal view of every page. Subsidiary works such as examining, on the monitor of the PC, images photographed by the digital camera, and storing them securely as image files, are carried out simultaneously with the main work. Here in the second of the two consecutive articles about the HUMI Project's photographing method of rare books, these works concerning image file handling that accompany the photographing work will be explained first. Then, the photographing method of special images different from that of digital facsimile, such as images of open pages and appearance of the book, will be presented. In the following part, the structure of the photographing set and the photographing process for handscrolls will be described.

Keywords：  HUMI Project / rare book / digitization / digital facsimile / photographing method / Naraehon / handscroll

Ikuko OBA
Elsevier Japan KK
Higashi Azabu 1-chome Bldg. 4F 1-9-15 Higashi Azabu, Minato-ku, Tokyo 106-0044 JAPAN

Abstract：Scopus is an abstract and citation database with over 15,000 titles from more than 4,000 publishers offering coverage of scientific, technical, medical and social science literature. Scopus continued its development to fulfill the needs of researchers to take the data and process in a variety of ways to get further insight. This paper discusses the characteristics of Scopus as an evaluation tool through introduction of Author Identifier, which helps to find articles from a specific author more accurately and easily, and Citation Tracker, which allows users to choose the articles they want to evaluate.

Keywords：  author search / citation analysis / Author Identifier / Citation Tracker / Scopus / A&I database / evaluation tool