This paper addresses a problem to identify risks in project plan documents written by project managers. Because some project managers express risks without negative words, risk identification method by negative expressions is not enough to identify them. In this paper, we propose a risk identification method by automatic acquisition of risk expressions. The proposed method acquires the risk expressions from existing project plan documents by using the frequency of appearance in the sentences including the risks. After the acquisition of the risk expressions, the proposed method extracts the sentence where the expression indicates the risk. From the extracted sentence, the proposed method identifies the risk by comparing the items of the checklist of risks. In particular, the proposed method identifies the items corresponding to each extracted sentence by using related words of each item. The experimental result shows that the proposed method identifies 10.7 known risks and 5.8 unknown risks on average.

This paper addresses a problem to reconfigure a bug thread hierarchy in bug tracking systems. The bug thread hierarchy consists of dozens of bug threads that have dependent relationships each other, and the hierarchy is reconfigured when a recent bug thread is posted. Not only bug threads that have dependent relationships with the recent bug threads but also their dependent targets or sources are similar to the recent bug thread. In our proposed method, similarities with dependent targets and dependent sources are weighted to the similarity with the bug thread, and the bug thread that has the highest weighted similarity is identified to have a dependent relationship with a recent bug thread. The dependent direction is decided by the similarity with dependent targets and sources. As a result of the experiment, the hierarchy of the bug threads is reconfigured at a correct rate of 81%.

This paper addresses the problem to identify the related group threads that has dependent relationships with recent bug threads. Because most of recent bug threads have no dependent relationships with group threads, basic approach based on similarity regards them as having dependent relationships wrongly. In this paper, we propose an identification method of related group threads by peak characteristics of similarities. The proposed method removes recent bug threads that have no dependent relationships by Support Vector Machine based on vectors representing peak characteristics of similarities between the recent bug thread and group threads. The application result shows that the precision rate is improved by 49% and the recall rate is kept 76% on average using the proposed method.

No abstract

No abstract