Search

In order to sell items of a genre that each user is not interested in, their latent interests of the genre have to be induced. Therefore, we focus on "Influencer," who is the person that has a large impact on a user's behaviors, and research a recommender system which advertises some items by utilizing influencer's remarks via social media. This paper investigates what kinds of remarks and remarkers would induce users'interests. As the result, we have revealed many findings. One is that "One of the factors that would induce the users' positive interests on the movie, is the positive reputation for the movie that is included in tweets." The other is that "If the user does not like the remarker, the user's negative interests for the items in the remarks of the person would be induced, even though the person has famousness.", INSPEC Accession Number: 18674306

To enable us to select the only scenes that we want to watch in a baseball video and personalize its highlights subvideo, we require an Automatic Baseball Video Tagging system that divides a baseball video into multiple sub-videos per at-bat scene automatically and also appends tag information relevant to at-bat scenes. The previous paper proposed several Tagging algorithms using ball-by-ball textual report and voice recognition. To improve our system, this paper introduces a more refined model for baseball games, and performs comparative experiments on models with regard to their recall, precision, and F-measure.

We experience or forecast various phenomena (e.g., rain, snow, earthquake) in the physical world, while we carry out various actions (e.g., blogging, searching, e-shopping) in the Web world. There have been many researches to mine the exploding Web world for knowledge about various phenomena and events in the physical world, and also Web services with the Web-mined knowledge have been made available for the public. However, there are few detailed investigations on how accurately Web-mined data reflect physical-world data. It must be socially-problematic to idolatrously utilize the Web-mined data in public Web services without ensuring their accuracy sufficiently. This paper introduces temporally-shifted Web Sensors with a temporal shift parameter δ to extract spatiotemporal numerical value about a physical phenomenon from Web documents searched by linguistic keyword(s) representing the physical phenomenon, and analyzes the spatiotemporal dependency of the temporal shift parameter δ with respect to their coefficient correlation with Japan Meteorological Agency's spatiotemporal statistics.

Various Web contents exist on the Internet in recent years, and are used by many users. But some people are so fascinated by and absorbed in Web contents that their everyday life is affected. Such a person yields to the temptation because of her/his weak will, even if s/he tries to control her/his Web access by self-control. Moreover, it is difficult for such a person to overcome the temptation because there are few tools to help her/his self-control. Then, this paper proposes a method that applies the filtering technology conventionally used for access restriction to self-control over Web access, and allows users to control their own Web access. Specifically, we create a prototype Web browser that has four functions to filter Web pages: a filtering based on their content, a filtering based on their ID (URL), a filtering based on their actions (buttons) such as purchasing and playing, and a time-based filtering.

In public spaces, there are a number of different contents such as visitors, physical information resources, and virtual information resources via their embedded output devices. Therefore, we might unexpectedly enter the public spaces that have our unauthorized contents and/or unwanted characteristics, i.e., they are not always secure and safe. To solve this problem, my previous work has introduced the concept of “Secure Spaces”, physical environments in which any visitor is protected from being pushed her unwanted information resources on and also any information resource is always protected from being accessed by its unauthorized visitors, and the model and architecture for space entry control and information access control based on their dynamically changing contents. Aiming to build more flexible Secure Spaces, this paper proposes an extended model for not only spatial entry control but also Ability-Based Expression Control according to how preferentially a virtual information resource should be outputted in a Secure Space where shared by visitors with perceptibility and understandability of the virtual information resource's content and expression.

Many researches on mining the Web, especially Social Networking Media such as web logs and microblogging sites which seem to store vast amounts of information about human societies, for knowledge about various phenomena and events in the physical world have been done actively, and Web applications with Web-mined knowledge have begun to be developed for the public. However, there is no detailed investigation on how accurately Web-mined data reflect real-world data. It must be problematic to idolatrously utilize the Web-mined data in public Web applications without ensuring their accuracy sufficiently. Therefore, this paper defines spatio-temporal Web Sensors by analyzing Twitter, Facebook, web logs, news sites, or the whole Web for a target natural phenomenon, and tries to validate the potential and reliability of the Web Sensors' spatio-temporal data by measuring the coefficient correlation with Japanese weather, earthquake, and influenza statistics per week by region as real-world data.

Many researches on mining the Web, especially CGM (Consumer Generated Media) such as Web logs, for knowledge about various phenomena and events in the physical world have been done actively, and Web services with the Web-mined knowledge have begun to be developed for the public. However, there is no detailed investigation on how accurately Web-mined data reflect real-world data. It must be problematic to idolatrously utilize the Web-mined data in public Web services without ensuring their accuracy sufficiently. Therefore, this paper defines the basic Web log Sensor with a neutral, positive, or negative description for a target phenomenon, and their linearly-combined Web log Sensors, and tries to validate the potential and reliability of these Web log Sensors' spatio-temporal data by measuring the correlation with weather (precipitation) and earthquake (maximum seismic intensity and number of felt quakes) statistics per day by region of Japan Meteorological Agency as real-world data.

Most researches on Image Retrieval (IR) have aimed at clearing away noisy images and allowing users to search only acceptable images for a target object specified by its object-name. We have become able to get enough acceptable images of a target object just by submitting its object-name to a conventional keyword-based Web image search engine. However, because the search results rarely include its uncommon images, we can often get only its common images and cannot easily get exhaustive knowledge about its appearance (look and feel). As next steps of IR, it is very important to discriminate between “Typical Images” and “Peculiar Images” in the acceptable images, and moreover, to collect many different kinds of peculiar images exhaustively. This paper proposes a method to search the Web for peculiar images by expanding or modifying a target object-name (as an original query) with its hyponyms based on hand-made concept hierarchies such as WordNet and Wikipedia.

This paper discusses PBL for first year software education, paying attention to aspects of holistic learning, and also proposes a new PLB course based on the discussion. The PBL course was given to high school students to confirm fundamental validity. The questionnaire method suggested that PBL might be fundamentally useful to learn holistically.

We have become able to get enough approvable images of a target object just by submitting its object-name to a conventional keyword-based Web image search engine. However, because the search results rarely include its uncommon images, we can often get only its common images and cannot easily get exhaustive knowledge about its appearance (look and feel). As next steps of image searches in the Web, it is very important to discriminate between “Typical Images” and “Peculiar Images” in the approvable images, and moreover, to collect many different kinds of peculiar images as exhaustively as possible. This paper proposes a novel method to precisely retrieve peculiar images of a target object by its typical/peculiar appearance descriptions (e.g., color-names) extracted from the Web and/or its typical/peculiar image features (e.g., color-features) converted from them, as a solution to the 1st next step of image retrievals in the Web.

Concept hierarchy knowledge, such as hyponymy and meronymy, is very important for various natural language processing systems. While WordNet and Wikipedia are being manually constructed and maintained as lexical ontologies, many researchers have tackled how to extract concept hierarchies from very large corpora of text documents such as the Web not manually but automatically. However, their methods are mostly based on lexico-syntactic patterns as not necessary but sufficient conditions of hyponymy and meronymy, so they can achieve high precision but low recall when using stricter patterns or they can achieve high recall but low precision when using looser patterns. Therefore, we need necessary conditions of hyponymy and meronymy to achieve high recall and not low precision. In this paper, not only "Property Inheritance'' from a target concept to its hyponyms but also "Property Aggregation'' from its hyponyms to the target concept is assumed to be necessary and sufficient conditions of hyponymy, and we propose a method to extract concept hierarchy knowledge from the Web based on property inheritance and property aggregation.

Mobile Web search engines as well as fixed ones become much more important for our Web access in the near future. However, mobile computing environments have troublesome limitations, as compared with traditional desktop computing environments. In order to solve the so-called "mismatched query problem" due to more shortness and ambiguity of mobile users' original queries, this paper proposes a novel method for context-aware query refinement based on their current geographic location and real-world contexts which are inferred from the place-names representing the location and the Web as a knowledge database

With the growing amount of information on the WWW and the improvement of mobile computing environments, mobile Web search engines will increase significance or more in the future. Because mobile devices have the restriction of output performance and we have little time for browsing information slowly while moving or doing some activities in the real world, it is necessary to refine the retrieval results in mobile computing environments better than in fixed ones. However, since a mobile user’s query is often shorter and more ambiguous than a fixed user’s query which is not enough to guess his/her information demand accurately, too many results might be retrieved by commonly used Web search engines. This paper proposes two novel methods for query modification based on real-world contexts of a mobile user, such as his/her geographic location and the objects surrounding him/her, aiming to enhance location-awareness, and moreover, context-awareness, to the existing location-free information retrieval systems.

We define "Secure Space" as a physical space in which any resource is always protected from its unauthorized users in terms of enforcing its authorization policies assuredly. Aiming to build such secure spaces, this paper proposes an architecture and a model for space entry control based on its dynamically changing contents, such as users, physical resources and virtual resources outputted by some embedded devices. We first describe the architecture and then formalize the model and mechanism for secure spaces.