MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641124687 A) filed by Rns Institute Of Technology, R R Nagar Post, Channasandra, Bengaluru.; Varun Brrncs; Ms. Jayashri Hassistant Professor; Ms. Purnima S Mittalkod Assistant Professor; Tarun Srncs; Vedanth Prncs; and Vivek Sdrncs on October 18, 2026, for Skill-Based Campus-Exclusive Peer Discovery And Real-Time Collaboration System Using Database- Layer Cosine Similarity And K-Nearest Neighbour Matching With Multi-Layer Skill Verification.
Inventors include Ms. Jayashri Hassistant Professor; Ms. Purnima S Mittalkod Assistant Professor; Tarun Srncs; Varun Brrncs; Vedanth Prncs; and Vivek Sdrncs.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: University campuses host thousands of students with diverse and complementary technical and non-technical skills, yet teams for hackathons, capstone projects and peer-learning activities are still formed informally through messaging-group broadcasts and word of mouth. As a result, capable students remain undiscovered and the teams that are formed are often poorly balanced. To address these challenges, the present invention provides a skill-based, campus-exclusive peer discovery and real- time collaboration system that identifies and connects compatible students automatically. The system restricts registration to verified institutional e-mail accounts and requires every new user to complete skill onboarding, in which each selected skill is assigned a proficiency level that is converted into a predetermined numerical weight. Every student is thereby represented by a non-empty weighted skill vector from the moment of registration, which eliminates the cold-start problem of recommender systems. A matching engine, implemented as a stored procedure executed inside the database engine itself, computes cosine similarity between the vector of a querying student and the vectors of all candidate students on demand and returns the K nearest neighbours, for example the top-20 matches, without any periodic batch computation. A dynamic scope-expansion strategy widens the candidate pool from branch and semester, to department, and then to the whole campus only when an insufficient number of matches is found, thereby countering popularity bias. The system further includes a four-layer skill verification pipeline comprising self- rating, a timed quiz, post-collaboration peer endorsement and linking of a code repository; a connection-gated real-time messaging channel that opens only after mutual acceptance of a connection request; and communities, events, project-showcase and faculty-expertise modules with role-based privileges for students, teachers and community heads. By combining weighted skill vectors, database-layer similarity computation, layered skill verification and institution-restricted access, the invention provides a fast, trustworthy and low-cost solution for structured team formation within a university campus.
Disclaimer: Curated by HT Syndication.
