Various applications have been realized on mobile computers such as smart phones and tablet computers. Because mobile computers have smaller monitors than conventional computers, strategies to develop user interfaces differ from conventional computer applications. For example, contents in a window are reduced or divided into multiple windows on mobile computers. To realize usable applications in this situation, usability evaluations are important. Although various usability evaluation methods for mobile computers have been proposed, few evaluate applications and identify problems automatically. Herein we propose a systematic usability evaluation method. In our method, operation histories by users are recorded and analyzed to identify steps with usability problems. Our method automatically analyzes usability problems, allowing usability evaluations in software development to be implemented easily and economically. As a case study, the operation histories were recorded and analyzed when 20 subjects operated an application on a tablet computer. Our method automatically identified many usability problems, confirming its effectiveness.
Since most production scheduling problems in Flexible Manufacturing System are NP complete problems, they are difficult to derive an optimum solutions. On actual production sites, it is required to make a schedule in a short time. In this research, we propose a look-ahead scheduling method with limiting candidates and depth to constrain the search space. In our experiments, look-ahead scheduling using two meta-heuristics was conducted and suitable strategy for the scheduling was shown. And our experiments show suitable strategy derives effective schedule in short time.
本研究では，CDCLソルバーの構成要素の１つであるリスタート戦略に焦点を当てる． リスタートは、探索状況が悪くなったときにその改善を目的として探索をやり直すヒューリスティクスであるが，本稿ではリスタートしてもしばしば探索状況が改善しないことを実証的に示す． そしてリスタート後の探索状況の改善を目的として，悪いリスタートの要因となる決定変数の頻出パターンが存在することを示す．