Area-based communication detector.

In this paper, we propose an area-based communication detection system. We aim to detect conversations with their attributes, such as communication length, communication members, relationship among members, and subjects. We think extracting correlations between conversations' attributes and office worker productivity is important for office worker productivity improvement. To evaluate our proposed system, we defined two kinds of evaluation metrics for communication length and members of conversations. Our proposed method improves detection accuracy by adapting detection algorithms to communication style. We conducted an experiment to evaluate our proposed method. The experimental result showed that our method achieved a 0.24 point higher F value than the baseline using only proximity data.