Neural network models have been actively applied to word segmentation, especially Chinese, because of the ability to minimize the effort in feature engineering. Typical segmentation models are categorized as characterbased, for conducting exact inference, or word-based, for utilizing word-level information. We propose a character-based model utilizing word information to leverage the advantages of both types of models. Our model learns the importance of multiple candidate words for a character on the basis of an attention mechanism, and makes use of it for segmentation decisions. The experimental results show that our model achieves better performance than the state-of-the-art models on both Japanese and Chinese benchmark datasets. 1

No abstract

No abstract

A study on designing task priority rule considering rework risk of system development project Abstract This paper proposes a methodology for designing task priority rule considering rework risk of system development project. In this paper, a process simulation considering reworks and the methodology for calculating optimal task priority rule by using genetic algorithm are developed. The process simulation considering reworks can be done by using the information of target system, rework probabilities, worker's skill and task priority rule. Proposed task priority rule consists of several dispatching rules that are related to each task. In developed process simulation, task priority rule can define which task is done preferentially in each time. In this paper, five dispatching rules are introduced for creating proposed task priority rule. By using developed process simulation, Monte-Carlo method and genetic algorithm, optimal task priority rule can be calculated from the viewpoint of average time required of target system development project. In this paper, proposed methodology was applied to a system development project. Results show that proposed task priority rule can get the lower average time required than other dispatching rule considering the structure of workflow and rework information.