Semantic Ideation Learning for Agent-Based E-Brainstorming

Brainstorming can assist organizations in generating creative ideas using teamwork and collaboration. However, the role of information technology in brainstorming is merely that of an assistant that passively supports the progression of brainstorming sessions rather than proactively engaging in the sessions. This paper integrates the unique association thinking of humans with an intelligent agent technique to devise an automated decision agent called the Semantic Ideation Learning Agent (SILA) that can represent a session participant who is actively participating in brainstorming. SILAs are grounded on the three association capabilities of human thinking (similarity, contiguity, and contrast). Furthermore, a Collective Brainstorming Decision System (CBDS) is built to construct an environment where SILAs can learn and share their knowledge with each other. Additionally, CBDS is integrated into an intelligent care project (iCare) for the purpose of innovated e-service recommendation. Preliminarily, evaluation results indicate that the proposed system advances e-brainstorming by crossing the three key boundaries of human ideation capability (understanding, cognition boundary, and endurance).