Jun Iio

Relationship between Reviews at App Store and the Categories for Software

The purpose of this study is making clear factors of what affects the contents of software reviews posted by customers at electronic commerce (e-Commerce) websites. Frequent referring to such reviews helps people to buy something. However, it is considered difficult to judge the quality of contents with reviews completely, because conditions for each software, such as, published data, categories, prices, are not equal. The result of this study has revealed that review styles are different with software categories, so that this paper concludes that consumers should pay attention to the bias of the reviews. Hereafter, it will be useful to investigate how to read reviews for not only consumers but also sellers, who want to know a value of software as well.

The Database on Near-Future Technologies for User Interface Design from SciFi Movies

Science fiction (SciFi) feature films offer viewers a glimpse into the future, revealing unique interfaces, social systems, and complex human relations. In this paper, we report a trial conducted by the Science Fiction Special Interest Group (SIG-SciFi) to gain insight into probable human-centered design (HCD) trends and the database of scenes collected to be used as good references for futuristic design. Characteristic scenes from the movie Minority Report, X-Men II, and The Island were analyzed. Our argument and analysis began with a top-down arrangement of scenes in descending order of importance. Then, extracted characteristic scenes were classified hierarchically while considering the worldview of the movie. As a result, suggestions were obtained pertaining to the direction of HCD in the near future. The results of this analysis are arranged into the database which can be immediately applied as a design tool.

Basic techniques in text mining using open-source tools

There are many text mining tools provided commercially and non-commercially. However, the elementary text-based analysis can be done with basic Unix commands, shell-scripts, and small program of scripting languages, instead of using such extensive software. This paper introduces the basic techniques for text mining, using combination of a set of standard commands, small code, and generic tools provided as the open-source software. The target of the analysis are sixty-seven articles written by one author in a relay column since 1998. Several text-based analyses reveals a trend of interest moved within about fifteen years. In addition, at the end of this paper, the results of text-based analysis are compared with that of non-text-based analysis and the efficiency of non-parametric analysis is discussed.

Document Management System Based on ISAD(G)

An efficient and effective utilization, the Open Source Software enables us to make a system which delivers low-cost Web services. Currently, there are many small and medium-sized archives and they do not have sufficient funds available to build up fully functional document management system (DMS). As an example of a DMS which can be used in such small and medium-sized archives, the DMS based on General International Standard Archival Description (ISAD(G)) has been constructed through the use of OSS components at a very low cost. This paper reports the overview of the DMS and the results of our project.

Business ecosystem for digital museums

There are a variety of research and development programs about museums using digital content and digital technologies (so-called “Digital Museum”). And some of those studies are trying to bring in technologies to existing museums. However, methods for collecting, applying, and sharing content at digital museums are totally different from those at traditional museums. Therefore, along with such studies, a new business ecosystem surrounding digital museums should be considered simultaneously. In this paper, an overview of network of people, goods, and capital surrounding digital museums is proposed. Providing a tangible example, it is discussed how to maintain self-sustainable operation at digital museums. With a success of this new business ecosystem, we can expect more technologies to be developed and flourish.

Newly-Added Functions for Video Search System Prototype with a Three-Level Hierarchy Model

This paper describes newly added functions onto ELVIDS, previously presented the three hierarchical video search system. We have developed this system for fostering scholarly use of videos over the three years. The development in the previous phase was mainly for giving functions for users such as retrieving videos, displaying search results, and playing back the videos. As the next phase, we considered adding functions for metadata registrars. In this paper, the background, sample dataset and a protocol for an authentication are firstly described in the earlier chapters. Then, the functions for metadata registration and the authentication including the methods and the interfaces are introduced. Finally, the conclusions and the further works are mentioned.

Anonymous Accessible Bulletin Board System with Location Based Access Control Mechanism

A bulletin board system (BBS) which allows anonymous submission has lower barriers to enter the discussions because everyone can post opinions without any responsibilities. Therefore, there are some risks that “a flaming” occurs, where the discussion is inundated with comments. To suppress emerging such situation, we developed a novel BBS with location-based access control mechanism. The BBS is anonymously accessible by default; however, the access permission to each chat room in the BBS depends on its geospatial information. That is, the chat rooms keep their geographical information when they are created, and users who can access each room must exist near the original location. In this paper, the overview of our system and its applications are described.

Lessons Learned from Data Preparation for Geographic Information Systems using Open Data

The use of geographic information systems (GISs) has become widespread in data-driven industries, and they are utilized to visualize various kinds of spatial data using mappings. In addition to the large amount of available open-source GIS software, various types of data (e.g., boundary data for administrative regions, numerical data for individual areas, and data representing the objects on a map) are provided by local governments as open data. However, in many cases, the data have been inadequately maintained. Thus, advance preparation is required to utilize the data effectively. This paper discusses the work required to utilize open data by considering the case studies of Hachioji-city, Tokyo, Japan.

ELVIDS: Video Search System Prototype with a Three-Level Hierarchy Model.

We propose a video search system with a three-level hierarchy model for searching videos from works, scenes, and shots. Videos are useful research materials; however, their scholarly use is very limited because finding the desired information is a time-consuming task for scholars. To provide solutions to this problem, we developed a system with functions to increase search efficiency. This paper mentions the reasons preventing the scholarly use of videos and describes the methods used to develop the system, focusing on database creation and the prototype’s interface.

Text-Based Video Scene Segmentation: A Novel Method to Determine Shot Boundaries

We propose a video data management system built on a hierarchical structure, in which each video is a work that is divided into scenes, and the scenes are further subdivided into shots to improve search efficiency. In our system, scene-level video segmentations are manually defined according to scene transitions. However, shot-level video segmentations within a scene should be defined automatically, because the numbers of shots is relatively large compared to that of scenes. Therefore, in this study, we developed a method to calculate shot segmentation automatically, based on information contained in closed captions, and we conducted an experiment using the method. This paper presents both a description of the novel segmentation method and the result of our experiment.