Thought Process Based Team Member Selection Using Contextual Sentiment Closeness
Hrishikesh Kulkarni · 2018
It is the teamwork that makes projects, companies and teams successful. Person alone is constrained by limitations. Even people with no shared thought processes and conflicting opinions cannot work together. Project undertaken by such teams lead to disastrous outcomes. Can two persons work together effectively? Can we determine and associate thought process to form effective groups? This paper focuses on determining thought process association of two persons to find out whether they can work together. This paper finds out closeness between responses by two individuals with reference to given context to determine whether they can work together in given scenario. The paper proposes contextual sentiment closeness (CSC) algorithm for the same. The algorithm further uses scenario pointer based reinforcement learning where team members collect thought process based rewards and penalties. Temporal Difference learning based approach calculates the maximum possible rewards and even suggest certain modifications required if any. This approach can help in selection of members for mission, projects and any critical activity. The selection based on responses compared with human selection on 200 samples and it gave over 81% accuracy.