Book review sites have been growing to which readers submit their own review. This study was aimed at investigating the characteristics of the online book reviews in comparison with those in newspapers. The sentences in the book reviews were analyzed based on the following points: target of evaluation, viewpoints, objectivity, and favorable evaluation or negative. The results were as follows: 1) 90 percent of the targets of evaluation were "the work itself". There was no difference between online reviews and those in newspapers. 2) 48 percent of viewpoints were "the writing style" in reviews in newspapers, while "plot" and "setting" were the main viewpoints in online reviews. 3) Most of the reviews in newspapers were objective and favorable, on the other hand online reviews vary in tone including subjective and negative evaluation.

View full abstract

In this research, we investigated whether a computer could estimate the best answer on a Q&A site. First, a best answer estimation experiment was carried out with human assessors. The data of Yahoo! Chiebukuro was used for the experiment; 50 questions extracted at random from four categories, viz., "Consultation of love," "Personal computer,""General knowledge," and "Politics," were used. The accuracy rate(precision) of the estimation by two assessors was 50% and 52% (random estimation: 34%) for "Consultation of love," 62% and 58% (random estimation: 38%) for "Personal computer," 54% and 56% (random estimation: 37%) for "General knowledge," and 56% and 60% (random estimation:35.8%) for "Politics." Next, the experimental results were analyzed, and the machine learning system with "Detailed","Evidence", and "Polite" in the feature as a factor to choose the best answer was constructed. The precision of the machine learning system exceeded the assessors' results in the "Personal computer"(67%) category, and it fell below the assessors' results in the "Consultation of love"(41%) category. In the "General knowledge" and "Politics" categories, the precision of the machine learning system was almost equal to the assessors' results.

View full abstract

Web archiving is to collect internet web pages, archive them, and make them available to the public. Wayback Machine of Internet Archive is a pioneer in this area, collecting web pages since 1996. Authors studied yearly changes of web pages of several Japanese organizations in Wayback Machine, and analyzed their public relation activities.

View full abstract

We studied development of the information system by the GMA (Global Model Architecture) concept and applied study 10 for several years. A basic function is an environmental offer to support thought supporting circumstances, group decision aiding environment and the agreement formation. We built the concept model of the remission of the triple handicap (ESET) of environment / society / the economy. And I proposed two major models. The first is a model keeping a language and the culture that only the human holds. The second is the human and a model keeping biology. Both have connection closely mutually, and it is the greatest problem of the human of the 21st century. The method to be realized of two major models is speed of the increase of the population and restraint of the degree of progress of the industrialization. The last purpose is remission of ESET. Two models have only Japan from the reason of the high quality nature of the Japanese culture. In this article, I assume it realization of two major models and COE, and information sends result. Therefore I introduced the wisdom of ancient people and these days. The two major models become minute by steady spiral activity. The effect will solve the greatest problem in the 21st century.

View full abstract

We could make the following thing clear by applying the conditions causing and solving contradictory to some social revolutions: 1. There are some positions a human can put himself in and normally he put himself in any position and coexist with other those who put themselves in the same and other positions. 2. Presently they begin to feel the contradictory between the new recognition and the present situation by recognizing as the same thing what he didn't recognize as the same thing. 3. Further they can solve the contradictory by understanding that the cause of contradictory is the differences of old positions bringing discrimination or unfairness because of what new recognition cannot apply to and the social institution supporting them. 4. Shortly the situation which needs the new recognition appears and when they understand that they can not resolve it because of their old positions, they begin to revolutionize their society irreversibly to the new position which their new recognition can apply to. 5. But the difference of their distances between the new recognition and their old positions cause prompting power for revolution and resisting power against revolution and it can become the origin of the social confusion. The significance which understand the revolution of our society in this way is to be able to know what stage our society is in when we encounter various revolutions and to prepare for.

View full abstract

To enhance one's ability surviving changing environment, one should continue to get and compose new knowledge that is necessary to adapt new situation in front, and at the instance one should try to evolve the connection between different knowledge in mind. To achieve the above information processing behavior effectively, it is the point that one could connect the newly got special-domain-knowledge with the existent special-domain-knowledge, and could recompose them to broader-domain-knowledge.

View full abstract

In order to scientifically investigate the cognitive processes of interpreting and evaluating texts, as a case study, this paper collects and numerically analyzes some literary reviews written by a famous critic. The results of co-occurrence analysis for evaluative vocabulary revealed 'newness', 'beauty' and 'depth' to be important. Moreover, by extracting individual names from the texts and detecting influential writers and philosophers, it was also possible to extract the background knowledge and philosophical characteristics that shape the standards of text evaluation.

View full abstract

Hoshi Shinichi, a Japanese writer, is known as one of the greatest short short-story writers in Japan, who wrote 718 works during his earlier years (1968-1973). Hoshi's works have often been criticized for their structural similarities. This paper investigates the narrative patterns of Hoshi's work and describes them in terms of 11 primitive units of narrative structure. Based on previous studies, these narrative units are established as the components of frequent narrative events within Hoshi's works. Employing these units and their combinations, the most basic narrative structures are extracted by conducting analyses of pattern distributions, factor analysis, and N-gram analysis.

View full abstract

In order to highlight conceptual features and to examine the relations to composers (musicians) evaluated by music critics, we focus on sensibility words and word phrases that describe composers and analyze them using a quantitative approach. We analyze the complete thirteen-volume set of "POLYPHONE", which is a highly respected journal of musical criticism within Japan. Each article was categorized according to seven groups based on seven most frequently occurring composers within all the articles; namely, Mozart, Beethoven, Wagner, Mahler, J.S.Bach, Debussy, and Schonberg, in descending order of frequency. We characterize the elements of sensibility linked to each composer concept by analyzing sensibility expressions, frequent sensibility words employed at the vocabulary level, and the usage features characteristic of each composer.

View full abstract

Similar to the research trends within various artistic realms, such as art, iterature, music, movies, and animation, in recent years, fiction studies are increasingly employing more sophisticated techniques of computational analysis. The aims of the present study are to quantitatively analyze the surface vocabulary of a literary text into the minimum components that reflect the writer's characteristic style and to identify style transitions. The target data consists of the full-length works of Haruki Murakami, as a representative of modern Japanese literature. The works are grouped in terms of writing styles through a clustering analysis of the textual vocabulary which was initially classified according to both word class and semantic categories. While the analysis results based on word-class categories yielded cluster that reflect a diachronic division, such as an "early trilogy", the results based on semantic categories yielded a cluster for the "Nezumi Tetrabiblos" that adds another work which is referred to as a sequel to the "early trilogy".

View full abstract

"Impact Factor" (IF) of Journal Citation Reports published by Thomson Reuters (or ISI) is applied to performance evaluation of researchers at research institutions, despite the warning by the publisher that IF should be used only for journal evaluations. But it takes several years to get meaningful citation counts for individual papers. The IF seems to be the only indicator instantly available for the annual evaluations, and it would be quite effective as an early approximation. The author proposes IDV: Impact Deviation Value, a normalized index comparable across fields, based on deviation values of IFs in each field. The actual IDVs are calculated and analyzed to demonstrate the effects with comparison with usual IFs. Special account for the situation of Japanese journals will also be added.

View full abstract

Our results based on analyses of co-authorship relations of academic publications between university and industry collaborations have shown that relations between university and industry are getting weaker, while collaborations with overseas are becoming stronger in Japan. This suggests that without taking account of collaborations with overseas it is no even more possible to explain national knowledge producing systems. In this study, we break down foreign sector into three regions of Americans, Asian areas and EU, to make further investigations on the formation of international research networks and time series changes of university-industry collaborations in Japan.

View full abstract

We have investigated genome relation, nanotechnology relation, information relation, etc. by keyword analysis about the trend and spread of the important research area of Science and Technology Basic Plan. Here, we compared the research task over about 20 years of National Institute for Environmental Studies for the trend of environmental field-related research with the research task of grants-in-aid for scientific research, and showed the result investigated about other relation and trends with an area of research by keyword analysis. Although it was an interim result, it turned out that the correlations of the classification field of National Institute for Environmental Studies and spatial relationship with the details field of grants-in-aid for scientific research are well shown by the used method.

View full abstract

In the last annual meeting, it was introduced that a database stored three kinds of risks data for sustainable supply of the metallic resources, environmental change, and toxic of materials was necessary to consider the sustainable supply of materials. As an example of the application of the database, the guide of the materials selection of the apparatus design used at high temperature based on the optimal environmental conformity was clarified. In this report, the data structure of the database to stored necessary information provided in the internet and to calculation TMR (total materials requirement), cost and risk of the sustainable supply of each metallic elements formed heat-resistant materials is discussed. Using the database, it is indicated that the guide of the materials selection of the apparatus design using at high temperature based on the possibility of the sustainable supply in those elements.

View full abstract

In this paper, we perform a case study of secondary batteries research and technology to develop a method of extracting commercialization gaps between science and technology. Scientific research tended to be basic, whereas patents focused on applied technology developed from the basic research. Of five science research fronts and five areas of technology, three scientific groups, Li(Mn,Co,Ni)O-, Li-Sn- and olivine-based batteries, corresponded to the largest set of patents relating to Li. These three academic outcomes are so recently published that these can be the candidates for industrial opportunities. Other than Li, the largest set of academic papers correlated semantically to the third largest set of patents. Although patents aimed physically such as covering the electrolytes more strictly, scientific researches aimed to polymerize the electrolytes used for secondary batteries, while. There also exists an opportunity for polymer electrolytes to solve the problems relating to electrolytes. Our results offer an intellectual basis for detecting possible opportunities for industry.

View full abstract

This paper aims at understanding a landscape of international collaboration among institutions in research fields of "Fuel cell" and "Photovoltaic" as green technology. This paper tries to visualize the structure and identify the top institutions. The innovation policy of green technology is proposed from the results.

View full abstract

In this study, we developed a platform for both expanding and utilizing name authority files of the National Diet Library. This software read JAPAN/MARC (A) format and enables many experts to complement lacked elements of name authority files through the Internet. It is also possible to add element which is not defined in MARC format and output name authority data to MADS/MODS format. With this platform, name authority files can be utilized for exploitation of advanced cultural information resources.

View full abstract

It is not possible to retrieve something that one wants deep in the mountains, using the full-text search system which we have developed for bibliographic data of documents in temple, since the system needs the access to the server. In this paper, we report a new system in stand-alone mode which works with an Ultra Mobile PC without communicating to any external computer. For constructing the system, the Windows operating system is employed instead of Linux, while the bibliographic data and the search engine software are unchanged. We implemented the interfaces for record registration and annotation as well as the features which the conventional system offered. The retrieval experiments made no difference between both search systems with regard to the retrieval results, and showed that the stand-alone system displayed results within two seconds after the user gave the search term.

View full abstract

Recently, as EUC/EUD has become more common, the user is ready gradually to participate in an information system. But there is almost no system design from this viewpoint in a database system for literature studies. Furthermore, a system designer almost tends to discuss only the technical aspects of a computer. In this report, we regard this point as a cause which a system does not activate. And I attempt the system design surrounding literary research activities including EUC/EUD.

View full abstract

This paper proposes an academic curriculum analyzing system which extracts the characteristics of them via the form of radar charts. This system treats syllabus data which constitute some curriculums, and calculates the similarity between the syllabus based on the occurrence frequency of technical terms. The radar chart view helps to find distinguishing features of the curriculums by visualizing and comparing the assignments of the syllabus to the clusters along various classification axes.

View full abstract

When using text and data mining to analyze existing data bases and web contents in order to develop service solutions aimed at users' knowledge-related needs, we must first ask in what way, and until what point, should we attempt to understand those needs? This paper summarizes results of a research project conducted within the framework of the "Needs Oriented Service Science and Engineering Research" (NEXER) program funded by the Japan Science and Technology Agency's Research Institute of Science and Technology for Society (JST-RISTEX). This research clarified social needs for which "knowledge structure" mapping and other visualizations based on judicial precedent and other legal information, would be useful. Lastly, it explores how related knowledge products and services could best be structured to benefit society.

View full abstract

When we integrate databases of Museums(M), Libraries(L), and Archives(A), there is a problem that the description rules of MLA are considerably different. Because the standard activity aiming for cooperation of MLA becomes active in recent years, we need to review description rules of MLA for the standardization of MLA. In this time, we compared three standards of IGMOI, ISBD(G), ISAD(G). In this paper, based on our comparison, we propose that which direction MLA cooperation should beresearched in the future.

View full abstract

Informationization of historical information is thought in a country and various organizations. It is difficult to think about the historical information at the same time. This informationization is not enough yet. Then, it thought about informationization of the festival of Japan (MATSURI). At first, I take definition of MATSURI and the trend of other historical information into consideration. Finally I attempt mapping in the Meta data and Dublin Core to make the database of MATSURI. Too special informationization becomes bad effect of enough profit use. Ultimately it needs to make basis and rules for use. Above all it needs each to make basis for historical information.

View full abstract