Takatoshi Ishii

Maximum Clique Algorithm and ItsApproximation for UniformTest Form Assembly

Educational assessments occasionally require uniform test forms for which each test form comprises a different set of items, but the forms meet equivalent test specifications (i.e., qualities indicated by test information functions based on item response theory). We propose two maximum clique algorithms (MCA) for uniform test form assembly. The proposed methods can assemble uniform test forms with allowance of overlapping items among uniform test forms. First, we propose an exact method that maximizes the number of uniform test forms from an item pool. However, the exact method presents computational cost problems. To relax those problems, we propose an approximate method that maximizes the number of uniform test forms asymptotically. Accordingly, the proposed methods can use the item pool more efficiently than traditional methods can. We demonstrate the efficiency of the proposed methods using simulated and actual data.

Development and evaluation of a self-learning support system for Patent Act suited to the current state of intellectual property education in engineering departments

In this paper, we propose a model of the problem-solving process for learning Patent Act, noting that legal texts can be expressed as logical structures. We also develop a learning support system for engineering students that uses this model.

Maximum Clique Algorithm for Uniform Test Forms Assembly

Educational assessments occasionally require “uniform test forms” for which each test form consists of a different set of items, but the forms meet equivalent test specifications (i.e., qualities indicated by test information functions based on item response theory). We propose two maximum clique algorithms (MCA) for uniform test forms assembly. The proposed methods can assemble uniform test forms with allowance of overlapping items among uniform test forms. First, we propose an exact method that maximizes the number of uniform test forms from an item pool. However, the exact method presents computational cost problems. To relax those problems, we propose an approximate method that maximizes the number of uniform test forms asymptotically. Accordingly, the proposed methods can use the item pool more efficiently than traditional methods can. We demonstrate the efficiency of the proposed methods using simulated and actual data.

A method for identifying customer orientations and requirements for product–service systems design

For manufacturers, developing product–service systems (PSSs) is getting more important because of the trends of servitisation and creating social value. A PSS is a social system where multiple actors mutually provide products and services. A PSS design, therefore, must take into account various actors as customers. However, existing methods provide an insufficient solution as to how various customers should be handled in an analysis to identify and accommodate various customer preferences and requirements. To tackle this issue, this article proposes a new method of identifying customers’ orientations and requirements for PSS design. The proposed method employs a combination of topic analysis, persona and scenario approaches. The effectiveness of the method is demonstrated with its application to an urban development case. Through the demonstration, its practical benefits are concluded as follows: consistent and logical results of requirement analysis and insights into a new market for manufacturers.

A Method for Supporting Customer Model Construction: Using a Topic Model for Public Service Design

For the design of public services, it is important to clarify service customers. For this purpose, various methods of customer modeling were proposed. Before constructing customer models, it is required to group customers and to characterize each customer group. However, the customer grouping based on some statistical barometers (e.g. age, sex, and job categories) may not reflect actual customer requirements for the service. This paper aims to propose a method for supporting customer grouping and characterizing without such statistical barometers. Finally, the proposed method is applied to an urban development case to demonstrate the effectiveness.

Face authentication for e-Learning using time series information

One problem in e-Learning is that cheating by impersonation is easy. We examined variations in facial images taken with a webcam during e-Learning with the aim of detecting impersonation. In previous work, we proposed an authentication method based on updating a learners facial images. This study examines weighted updating and the use of only the maximum similarity calculated instantaneously during the e-Learning session. The findings indicate that these proposed methods improve authentication accuracy.

Algorithm for Uniform Test Assembly Using a Maximum Clique Problem and Integer Programming

Educational assessments occasionally require “uniform test forms” for which each test form consists of a different set of items, but the forms meet equivalent test specifications (i.e., qualities indicated by test information functions based on item response theory). For uniform test assembly, one of most important issues is to increase the number of assembled tests. This study proposes a new algorithm, RIPMCP, to improve the number of assembled tests. RIPMCP applies a maximum clique algorithm and integer programming for assembling uniform tests. RIPMCP requires less computational space resources, thus, the proposal can assemble a greater number of tests than the previous methods on the same computational environment. Finally, we demonstrate the advantage of the proposal using simulated and actual data.

A framework to search better series of lecture improvement

To increase learning quality, instructors and learners must make efforts toward improving their lectures. As a start, instructors and learners need to share context and know-how, and cooperate for lecture improvements. We propose a framework for capturing learning states and context, and for better searches for lecture improvements. Specifically, this framework records and analyzes performed actions in lectures that contains efforts toward improvements, their context, and the result of the actions as described in lecture evaluations.

Item bank to estimate the answers of class evaluation questionnaire

In Japan, student class-evaluation questionnaires are part of faculty development at universities. However, it is difficult to improve lectures from the results of a questionnaire containing only a few items. Here, we attempt to suggest improvements from a small questionnaire. For achieving this, we propose an item bank that can estimate the answers to questions that were not asked. Such an item bank could provide information to help instructors improve. More specifically, we assume that the relations between questionnaire items are probabilistic causal relations. Then, we discuss a way to provide feedback to instructors by simulation.

A Topic Model for Clustering Learners Based on Contents in Educational Counseling

For improving the quality of education, we need to analyze the interaction among the learners and teachers. For example, we empirically know that an agreement among the learner and teacher on the point of learning motivation makes good lecture. For this purpose, this paper aim to characterize the interactions based on the contents in the interactions. This paper employs a topic model for characterizing the interactions. Topic model is a method for estimating topic (theme or subject) in documents and clustering the documents based on estimated topics. By using topic model, this paper analyzes contents in actual educational counseling.