SIdeal, System and Method for Attribute Weight Induction in a Multiple Recruiter Setting Exploiting Public Goods Games Framework
IN820170195US01 | US 2017 (Patent Pending)

Sudhanshu Singh, Gyana Ranjan Parija, Ritwik Chaudhuri, Manu Kuchhal, Sarthak Ahuja, Ritwik Chaudhari, Manish Kataria

Here we disclose a public goods games (PGG) based system to define and refine relative weights of various attributes defining the profile. The system consists of a SIdeal agent corresponding to each recruiter with a preferred weightage for various attributes. These agents are engaged in a sequential gameplay where they modify and update their preferences to eventually converge on a consensus.

System and Method to produce Generalized Representation of Job Description Documents and Calculate Similarity using the Representation in Recruitment Domain
P201704553US01 | US 2017 (Patent Pending)

Joydeep Mondal, Sudhanshu Singh, Sarthak Ahuja, John Medicke, George David, Amanda Klabzuba

We propose a system and methods to produce a action-object-attribute triplet representation of a job description document. In addition to this we also propose to calculate similarity between two job description documents using such representations.

App-lause - VR based Cultural Audience Simulation for Immersive Rehearsals
XX | US 2017 (Awaiting File)

Sarthak Ahuja, Kushal Mukherjee, Joydeep Mondal, Sudhanshu Singh

We disclose a system and associated methods for simulating a relevant audience for an activity in context and provide feedback in the form of stimuli on a person’s performance sensed by input sensors based on cultural metrics. These cultural aspects are assumed to be specific to the place of performance and are mined from online resources specific to the region.

Cogniculture based Eco-System for Multi-Viewer Smart TVs
XX | US 2017 (Awaiting File)

Sarthak Ahuja, Kushal Mukherjee, Joydeep Mondal, Sudhanshu Singh

The main novelty of our system is a decentralized agent based architecture for the complete TV show recommendation and watching experience in both a multi-user and single-user setting. Each agent is allowed to have it’s own unique method to compute recommendations for it’s assigned user. We detail an ecosystem and define it’s rules, where these agents can autonomously collaborate and generate these recommendations.