Bifurcated Code Summary Clustering for Redundant Code Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing team-based computer programming practices result in redundant code development due to the lack of a common taxonomy and lexicography for organizing and assembling native code scripts, and the application of artificial intelligence is hindered by the need for large amounts of high-quality training data and the inability to verify results accurately.

Innovation Solution

A bifurcated model architecture comprising a first model to generate code summaries and a second model to compare and categorize them into clusters, using a large language model (LLM) and natural language processing (NLP) to overcome these challenges.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If artificial intelligence models are used to organize and assemble native code scripts, then code organization and assembly can be automated, but the need for large amounts of high-quality training data creates complexity and time consumption

Engineering Contradiction:
Improvecode organization automationVSAvoiddata preparation complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary representation layer that translates native code scripts into a standardized intermediate format before feeding them to AI models. This intermediary layer acts as a mediator that simplifies the data preparation process by providing a common taxonomy and lexicography, thereby reducing the complexity of training data requirements while enabling automated code organization and assembly.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Extent of automation

If artificial intelligence models are used to generate code summaries, then code processing can be automated, but the inability to verify results accurately creates technical problems

Engineering Contradiction:
Improvecode processing automationVSAvoidresult verification accuracy
Core Design Contradiction:
Extent of automationVSReliability

Solution Approach 1:

The patent implements feedback mechanisms where generated code summaries are automatically verified against the original code scripts and existing codebases. The system provides feedback loops that allow continuous improvement of model outputs, enabling automated code processing while maintaining accurate verification through comparative analysis and validation protocols.

Inventive Principle:
Principle #23Feedback

3Productivity

If teams work in distributed fashion to generate code, then productivity can increase, but redundant code development occurs due to lack of common taxonomy

Engineering Contradiction:
Improvedevelopment speedVSAvoidredundant code
Core Design Contradiction:
ProductivityVSLoss of substance

Solution Approach 1:

The patent establishes a universal code representation framework that serves multiple functions: organizing code, generating summaries, identifying similarities, and facilitating collaboration. This multi-functional system enables distributed teams to work efficiently by providing a common language and taxonomy, thereby increasing productivity while eliminating redundant code development through standardized classification and indexing.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS20260111221A1Systems and methods for a bifurcated model architecture for generating concise, natural language summaries of blocks of code
Publication Date: 2026.04.23 CAPITAL ONE SERVICES LLC
  • US20260111221A1 patent drawing
  • US20260111221A1 patent drawing
  • US20260111221A1 patent drawing

AI summary

Systems and methods are described for generating code summaries using bifurcated model architectures. The bifurcated model architecture may comprise a first model that generates code summaries based on native code script, and the second model may compare, de-duplicate, and/or categorize the code summaries into clusters that perform the same or similar functions. That is, the systems and methods may comprise a first model (e.g., a large language model) to predict new content (e.g., a code summary for a given native script code). The outputs of the first model are then inputted into a natural language processing (NLP) model to compare the code summaries. For example, the first model of the bifurcated architecture performs a crucial pre-processing step that prevents the NLP model from generating cluster descriptions that are specific to semantic structures, programming languages, and/or code formatting.