Design of an Intelligent Agent for Stimulating Brainstorming
Chun-Hsiang Wang, Tsai‐Yen Li · 2018
In recent years, brainstorming has gradually become a mainstream approach used by groups of people to collect ideas before making decisions. This approach is useful in handling the predicament of lacking ideas on specific topics for individual members, and breaking through the limitations of their own experiences. However, in the process of group brainstorming, there exists some common problems such as the lack of comprehensive discussion contents due to the limited experiences recalled by the members, the stagnation in the progress, and so on. Therefore, in this paper, we propose the design of an intelligent agent system that plays the vital role of facilitator to participate in the brainstorming process; in other words, our system can discuss a designated topic on an on-line chatroom with other members in a brainstorming discussion. This system collects textual data and establishes the knowledge model in a specific domain with machine learning methods prior to the brainstorming. Then, in the brainstorming process, it attempts to conjecture the topic of the current dialogues, determine the progress, and generate responses to the brainstorming chatroom in order to come up with diverse ideas or extend the present topic. The goal of this system is to increase the diversity and profundity of discussions, and make the whole process smooth and fruitful. The results of our experiments show that our intelligent agent is effective in helping on-line brainstorming, especially idea generations.