田村 恭久

<p>This paper focused on students' peer assessment and analyzed evaluation behavior using log data from the evaluation process. Peer assessments by students are problematic in terms of reliability and validity. Many previous studies have discussed the reliability and validity of peer assessments using the evaluation scores of peer assessments. In contrast, the evaluation process, that is, who, when, and which items were evaluated, and in what order, has not been studied. The purpose of this research is to answer this research question, "How do students evaluate each other during peer assessments?" To achieve this purpose, we break the research question into three sub-questions: (1) Is it worth focusing on evaluation behaviors? (2) How can we visualize evaluation behaviors? and (3) What are the characteristic evaluation behaviors? First, we investigated similar studies and considered their relationship with our proposal. Consequently, it became clear that in social science, there have been many studies that acquire process data for answering surveys and questionnaires. Therefore, also in peer assessment research, there is a possibility to gain a better understanding of how students participate in peer assessment by focusing on the process log. Second, we developed a web-based form as the peer assessment tool to detect students' evaluation process data. Using this form, we conducted an experiment to acquire evaluation process data from a class in which assessments were made to visualize evaluation behavior. Finally, we extracted and discussed characteristic evaluation behaviors. Consequently, we identif ied characteristic evaluation behaviors such as "evaluated in a short time." We expect these results will lead to a clarif ication of the causes of low reliability and validity in students' peer assessments that have been dif f icult to assess through conventional methods based only on the peer assessment score. Specif ically, the causes are, for example, "the students were not evaluating seriously," "they evaluated seriously but could not evaluate properly because of a lack of evaluation ability," or "there was room for improvement in the form design or lesson conditions."</p>

The authors verified learners' learning styles with use of learning style questionnaires and page flip history of materials. First, page-flip detection function of material slides on learners' PCs was developed. This function also transfers the result to a dedicated LRS (Learning Record Store) server automatically. From this fine-grained learning activity log, student's learning styles were classified. Second, the authors estimated learners' learning styles with use of widely used learning style questionnaires. The authors verified these two types of learning styles for 100 subjects (research participant) in real 4 units of 2 classes. As a result, a student who "rushes up ladder" during a certain time period was significantly relevant to "Global" learning style tendency rather than "Sequential" one.

This paper introduces an automatic analysis and feedback system for amateur baseball pitching learners with use of Microsoft Kinect. Compared with conventional 3-D motion capture systems, Kinect has advantages of reasonable cost and easiness of application system development for physical exercises. The authors developed a similar system for flying disc throw, which is disclosed in Tamura, Yamaoka et al. (2013) and Tamura, Uehara et al. (2013). In this paper a target motion is moved to baseball pitching, so focused body parts and judgment criteria were changed, although utilizing the same Kinect platform. The proposing system acquires postures and motions of amateur baseball pitchers, and judges them in 2 criteria: (1) maximum angle of elbow, and (2) hand twist, which were decided by a training expert. It also displays feedback messages to improve their motions. As a result, novice testees of the target group showed significant improvement of their pitching motions.

This paper shows typical functions required on electronic textbooks to be adopted in various countries, trial implementation of these functions, and their verification of representation and behavior in six e-book platforms. First, the authors extracted the functions by checking preceding researches and investigations. The result contains 52 items, including 7 categories of “authentication”, “copyright”, “contents representation”, “related information”, “additional information by learners”, “learning support”, and “restriction of contents and platforms”. Next, these functions were implemented with ePub3 format e-books. ePub3 is going to be one of de jure standard format of e-books. Also, contents representation and behavior of the result ePub e-books were verified on 6 types of e-book platforms on operating systems of iOS, Android, Windows, and Kindle Linux. The verification result shows that iBooks3 on iOS worked best among the 6 e-book readers. Also it shows that many e-book readers support only representation of ePub2 files, while HTML5, JavaScript, and Web API are not supported yet.

This paper proposes a three-dimensional motion capture and feedback system of flying disc throwing action learners with use of Kinect device. Rather than conventional 3-D motion capture system, Kinect has advantages of cost merit, easy system development and operation. A novice learner of flying disc is trained to keep arm movement in steady height, to twist the waist, and to stretch the elbow according to the waist angle. The proposing system captures learners’ body movement, checks their skeleton positions in pre-motion / motion / post-motion in several ways, and displays feedback messages to refine their actions.

This paper shows typical functions required on electronic textbooks to be adopted in various countries, trial implementation of these functions, and their verification of representation and behavior in six e-book platforms. First, the authors extracted the functions by checking preceding researches and investigations. The result contains 52 items, including 7 categories of "authentication", "copyright", "contents representation", "related information", "additional information by learners", "learning support", and "restriction of contents and platforms". Next, these functions were implemented with ePub3 format e-books. ePub3 is going to be one of de jure standard format of e-books. Also, contents representation and behavior of the result ePub e-books were verified on 6 types of e-book platforms on operating systems of iOS, Android, Windows, and Kindle Linux. The verification result shows that iBooks3 on iOS worked best among the 6 e-book readers. Also it shows that many e-book readers support only representation of ePub2 files, while HTML5, JavaScript, and Web API are not supported yet. © 2013 The Authors.

This paper proposes a three-dimensional motion capture and feedback system for flying disc learners with use of Kinect device. Compared with conventional 3-D motion capture systems, Kinect has advantages of cost merit, easy system development and operation. The proposing system captures learners' body movement, checks their skeleton positions in pre-motion / motion / post-motion in several aspects, and displays feedback messages to improve their motions. A novice learner of flying disc is trained to keep arm movement in steady height, to twist the waist, and to stretch the elbow according to the waist angle. A result of controlled experiment shows that the proposing system is effective for beginners to improve their movements. (C) 2013 The Authors. Published by Elsevier B.V.

This paper proposes a function to restrict undesired utterance roles on CSCL, in order to accelerate better discussion. The authors found that CSCL discussion often falls into stagnated and distracted situation when continuing utterances have same utterance role. To avoid this situation, the proposing system monitors the utterance roles attached by learners, finds continuity of utterances with the same role, and prevents to use the target role for the next utterance upload. Result of controlled experiment shows that the proposing function works well, and learners tend to refer more preceding utterances.

This paper proposes an automatic analysis and operating functions of CSCL utterance logs. First, the authors specified a data scheme called COLCOX and relevant XML bindings in order to represent CSCL utterance with structure and meanings. Next, an automatic analysis method was proposed to resolve elliptical phenomena in CSCL discussions with use of case analysis for spoken words. Further, an automatic exchange function was developed between LMS. An open source LMS of Moodle was utilised and expanded to install these functions. With use of the proposing functions, a learner group is able to search related utterances of various standpoints. Copyright © 2008 Inderscience Enterprises Ltd.

The authors specified XML bindings for CSCL data interchange, and developed an automatic CSCL data exchange function among LMSs. Conventional CSCL activities are processed without exchanging any information with other learner groups, even they are discussing the same topic. With use of the proposing function, a leamer group is able to acquire basic information, related references and various standpoints for the given topic during the discussion or in advance. In order to execute XML based inter-LMS data transmission, XML bindings named COL-COX was specified. Also, an open source LMS of Moodle was utilized and expanded to install the proposing function to transmit CSCL data. With use of the proposing function, a leamer group is able to get scaffolding information and knowledge on CSCL activities.

The authors developed a CSCL environment to utilize "Six Thinking Hats". This environment is an extension of online discussion room, and extended to assign "hat" to each learner, facilitate to think based on his "hat",and visualize which hat is on. With use of this environment, an instructor is ableto introduce, facilitate, and train students "Six Thinking Hats" method in onlinelearning environment with less instruction workload compared with face-to-face learning environment. The authors also evaluated the proposingenvironment, and found significant difference of number and quality of ideasfor given problems between "Six Thinking Hats" environment and simple online discussion environment.

ドメイン分析・モデリングは,対象システム自身が本来もつ各種の性質や開発上の多様な知識を十分に分析し認識して組織化し,システムの開発に有効な,共通の対象領域(ドメイン)に属する,用語,問題の捉え方,システムの構造,システムの作り方等の,固有な概念構造を得るプロセスである.この概念構造をドメインモデルと呼ぶ.このドメインモデルを用いて,複雑で大規模な実際のシステム開発での生産性の向上と再利用の促進を図ろうとしている.得られたドメインモデルは,そのドメインに属するシステムを繰り返して開発する際の分類モデル,ひな型モデル,標準モデル,ハンドブック等として使われる.本概説では,ドメインの意味,ドメイン分析・モデリングの目的とプロセス,ドメインモデルの構成要素・記述手段・種類を述べ,また,ソフトウェアの再利用,ドメイン工学,要求工学,設計,リバースエンジニアリング等の観点からソフトウェア工学との関係を述べる.

This paper proposes a bottleneck improvement method on a new type of Queueing Network with Shared Stores (QNSS) using qualitative reasoning. Based on heuristics and knowledges obtained from evaluation experts, the authors have developed a new "qualitative reasoning" -based expert system, named BDES-SS (Bottleneck Diagnosis Expert System for QNSS), which can identify the bottlenecks, analyze the sources, and provide qualitative improvement plan option for QNSS. QNSS has the different types of bottlenecks in comparison with the ordinary QN. Primary types of bottlenecks in QNSS are "bottleneck Servers" with high utilization rates, "shared stores in overflow bottleneks", and "shared stores in underflow bottlenecks". In case of bottleneck improvement of QNSS with more than one bottlenecks, improvement Plans cause inconsistency one another. In order to improve all bottlenecks, BDES-SS first enumerates improvement plans for each of bottlenecks. Then, it selects a feasible Combination of the plans without inconsistency between bottlenecks. For this combination, the authors propose an approach by qualitative influence Propagation on the skeleton network which is a reduced from of given QNSS.

対象システムが固有に待つ性質に着目して、複雑で大規模な実際のシステム開発における生産性の向上と再利用の促進を目的とする「ドメインモデリング/分析」について解説する。

The authors present the Triadic Domain Model (TDM)-based system development environment. The main purpose is to integrate system specification, system design and system implementation of software system using domain specific properties. TDM consists of Domain Problem Model, Domain Product Model and Domain Process Model.

In this paper the authors propose a performance-oriented design/re-design methodology for pipelined computer architecture with use of heuristic logic modification. Hardware data flow of the CPU (Central Processing Unit) can be modeled as the network. However, no mathematical formalization and analysis of this network are available, because usually many entities of the network are tightly connected and often form loops. The proposing method executes a performance simulation, identify the bottleneck or an entity to be improved, and modify the network with use of modification rules. This method is effective for both architecture design at the first time and architecture re-design, that designs another type of CPU with same software-hardware interface but with different hardware technology and implementation, to modify the implementation of the former CPU design.

A software prototyping environment called TransObj (TRANSaction and OBJect) is used for designing real-time Transaction-based Concurrent Software Systems (TCSS). In a TCSS design process, a software designer should perform both functional design and performance design. The designer should change his design view from a transaction-based paradigm to an object-based paradigm during the TCSS design process. Recognition of re-entrant functional objects and serially reusable functional objects in the TCSS should be required. TransObj includes the Stepwise Prototyping Method (SPM), and two SPM-based tools: Prolog-based TransObj (P-TransObj) and GPSS-based TransObj (G-TransObj). SPM enables the designer to advance both functional design and performance design for the TCSS prototype as controling the change of design view paradigms. P-TransObj mainly checks the prototype in a microscopic view on a personal computer. G-TransObj mainly checks the same prototype with a longer time span on a large-scale computer.