Dynamic Topic Definition Generator for Text Summarization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current systems lack the ability to quickly and effectively summarize high volumes of information from large datasets, such as customer complaints and social media, leading to delays in identifying trends and issues, which can result in lost revenue and reduced usefulness of trend analysis.

Innovation Solution

A system that uses machine learning and natural language processing to generate topic sentences from a corpus of documents, summarizing topics and their contexts by identifying frequently recited terms, removing irrelevant text, and creating sentences based on probability data, allowing for rapid interpretation and action.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If machine learning and natural language processing are used to generate topic sentences from large document corpora, then the speed of identifying trends and issues is improved, but the complexity of the system increases

Engineering Contradiction:
Improvespeed of identifying trends and issuesVSAvoidsystem complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The system segments the document corpus into manageable chunks and processes them through multiple specialized components: a topic model generator that identifies topics, a sentence generator that creates topic sentences, and a summarization model that produces condensed summaries. This segmentation allows complex data to be processed systematically and efficiently.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediary components including a topic model that acts as a mediator between raw documents and final summaries, and a sentence generator that bridges the gap between topic identification and summary production. These intermediaries simplify the overall system architecture by breaking down complex processing tasks into manageable stages.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Productivity

If automated topic sentence generation is implemented, then productivity in summarizing high volumes of information is improved, but the precision and contextual accuracy of topic representation may deteriorate

Engineering Contradiction:
Improveproductivity in summarizing informationVSAvoidprecision of topic representation
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system incorporates feedback mechanisms where generated topic sentences and summaries are evaluated against the original documents and user preferences. This feedback loop allows the model to iteratively improve its accuracy in representing topics while maintaining high processing throughput.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent employs parameter changes by adjusting topic model sensitivity, sentence generation probability thresholds, and summarization length parameters to optimize both productivity and precision. These parameters can be dynamically tuned based on document characteristics and user requirements.

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If irrelevant text is removed and only frequently recited terms are used, then the clarity and usefulness of topic sentences is improved, but the loss of information from rare but important terms increases

Engineering Contradiction:
Improveclarity of topic sentencesVSAvoidinformation loss from rare terms
Core Design Contradiction:
Ease of operationVSLoss of information

Solution Approach 1:

The system applies partial action by selectively processing only the most relevant portions of text—those containing frequently recited terms—while using contextual analysis to infer meaning from less common terms. This approach maintains clarity by focusing on dominant patterns while preserving important niche information through intelligent inference.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent uses copying by generating synthetic topic sentences and summaries that replicate the structure and style of high-quality manual summaries. These generated copies capture the essence of rare terms through pattern recognition and contextual modeling, preserving their informational value while maintaining readability.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS20240281611A1Dynamic topic definition generator
Publication Date: 2024.08.22 WELLS FARGO BANK NA
  • US20240281611A1 patent drawing
  • US20240281611A1 patent drawing
  • US20240281611A1 patent drawing

AI summary

Disclosed in some examples are methods, systems, and machine readable mediums which provide summaries of topics determined within a corpus of documents. These summaries may be used by customer service associates, analysts, or other users to quickly determine both topics discussed and contexts of those topics over a large corpus of text. For example, a corpus of documents may be related to customer complaints and the topics may be summarized to produce summaries such as “credit report update due to stolen identity.” These summarizations may be used to efficiently spot trends and issues.