International Business Machines Corporation
Empathy fostering based on behavioral pattern mismatch

Abstract:

A cognitive system collects online behaviors of a user and an affinity group of users who are related (e.g. by relationship, or behavioral similarities) to the user. A knowledge base of behavior and sentiment patterns is produced and maintained. If real-time data for the user shifts in behavior and/or sentiment and significantly deviates from established patterns, the system looks for a similar behavior and/or sentiment pattern shift among members of the affinity group. If the affinity group patterns shift in a manner similar to the first user's pattern shift, the cognitive system, in response, updates the knowledge base with information related to the shift, thereby adding knowledge to the long-term patterns. If the cognitive system finds that the user's behavior and/or sentiment pattern shift differs significantly from the affinity group, the system generates an empathy fostering alert message and sends it to one or more recipients.