1. DST Sponsered Ongoing Project

"Development of Distributed Deep Learning based Dynamic Causality Analysis Model for Predicting Epilepsy and Simulation of Epileptic Brain" Funded by Dept of Science & Technology under Cognitive Science Research Initiative (CSRI) for 2020-22

PI : Dr. Vishwambhar Pathak

Co PI : Dr. Vivek Gaur

2. DST Sponsered Completed Project 

“ Soft modeling of group dynamics : A cognitive psychology perspective” has been sanctioned by Department of Science & Technology, Ministry of Science & Technology, Government of India. Duration of project : 2009-2011 (Two Years)

Objectives:

• To analyze group dynamics, behavior and its impact to the performance and movement of a group under different paradigms.
• To  study programmer’s mental model behind the software development and selection of group and group leader .

The extension of the outcome of the proposed project could be ported to other similar group of problems and construction of forecasting model involving a group. The Group may be the group of crowd affected in a war situation or by natural disaster moving from one region to another.

Methods Adopted:

The proposal has considered a simple case study of software development scenario, where there are mental involvements among the members in the context of software development. Which is analyzed and key players are identified.

The work is divided into following Steps:

1. Development of Message Board
2. Scenario Graph Generation.
3. Community Detection
4. Discovering Communities
5. Identifying key Players

Experimental output :  

Experiment 1: Discovering Communities or teams through Bibliometric approach and edge betweenness successively

Experiment 2: Community Detection and Key Player Identification in terrorist network

Experiment 3: Identifying Key Players or Team Leaders in a Student’s discussion network (while doing software development activity) through Neighbourhood Connectivity measure

Figure 1: Student’s discussion network during software development activity.

Figure 1  is the snapshot of the network of Students while they are participating in Software development activity. Every node is a student and every node number is the student class roll number. In figure 2, Detected teams are in different colors and identified team leaders are indicated by a circle around the node.

Figure 2: Identified candidate key players or team leaders (10, 18) in teams (communities) detected in the Student’s discussion network during software development activity

Conclusion :

For the application based project development,  Team and team leaders were selected using the designed algorithm and the outcome is very satisfactory  as the complete project is completed within time limit and the quality of the project is up to the mark. The designed message board and scenario graph approach is efficient for community detection and team forming applications