MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202641084026 A) filed by Amrita Vishwa Vidyapeetham on July 08, 2026, for A System And Method For Generating Natural-Language Summaries Of Source Code.
Inventors include Viswanathan, Sunitha Edacheril; and Koshy, Eric Binu.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: The present invention relates to a system and method for generating natural-language summaries of source code. The system may include a structural analysis module configured to receive source code and generate a unified multi-graph representation encoding structural and behavioral properties derived from at least two distinct program analysis representations, including a control flow graph, a data dependency graph, a function call graph, and an abstract syntax tree. An identifier anonymization module may replace user- defined identifiers with semantically neutral placeholder labels prior to any language model inference, retaining a reversible mapping for post-inference restoration of original identifiers in the generated summary. A prompt construction module may assemble a prompt comprising the anonymized source code and a representation of the unified multi-graph representation, which a language model inference module may receive to generate the natural-language summary. Figure 1 is the representative figure.
Disclaimer: Curated by HT Syndication.
