Topic Tracking Platform Using Machine Learning for Emerging Trends

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for processing large volumes of feedback data, such as employee and client complaints, struggle to identify emerging topic trends due to their reliance on manual searches and fixed taxonomies, which limits the sensitivity and specificity of complaint search results and can escalate issues.

Innovation Solution

A topic management platform utilizing a machine-learning model that processes a corpus to identify topics hierarchically organized components, providing a topic map with contributions from each component, and a visualizer to analyze and display information at different hierarchical levels, enabling quick flagging of emerging trends and refining the accuracy of topic tracking.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual searching with fixed taxonomy is used to process feedback data, then the search process is simple and controllable, but emerging topic trends cannot be identified and measurement precision deteriorates

Engineering Contradiction:

Improvetopic tracking sensitivityVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces manual searching and fixed taxonomy systems with a machine-learning model that automatically processes feedback data. The MLM learns patterns and identifies emerging topics without relying on pre-defined taxonomies, thereby improving measurement precision while the automation reduces the need for complex manual search procedures

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system dynamically adjusts topic identification parameters by training the machine-learning model on evolving data patterns. Instead of using static fixed taxonomies, the model adapts its parameters to recognize emerging trends, improving sensitivity while the automated parameter adjustment manages system complexity

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If fixed taxonomy searching is used, then the search method is straightforward and easy to operate, but topic identification accuracy deteriorates

Engineering Contradiction:

Improvecomplaint search result accuracyVSAvoidsearch operation simplicity

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The patent replaces straightforward but inaccurate fixed taxonomy searching with a machine-learning-based automated system. The MLM automatically processes feedback data and identifies topics with higher accuracy, while the automated nature of the system maintains ease of operation by eliminating manual search requirements

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The machine-learning model performs self-learning and automatic topic identification without requiring manual search operations. The system serves itself by automatically processing feedback data, determining topic hierarchies, and generating results, thereby improving accuracy while maintaining operational simplicity

Inventive Principle:

Principle #25Self-service

3Loss of information

If automated processes with fixed taxonomy are implemented, then processing efficiency is improved, but emerging trends are not identified and information loss occurs

Engineering Contradiction:

Improveemerging trend identificationVSAvoidautomated process complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent replaces automated fixed-taxonomy processes with a machine-learning model that can identify emerging trends. The MLM automatically learns from data patterns without being constrained by pre-defined taxonomies, capturing emerging information that fixed systems would miss, while the automation manages processing complexity

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system transitions from static fixed taxonomy to dynamic machine-learning-based topic identification. The MLM adapts to changing data patterns and emerging trends in real-time, ensuring information is captured while the dynamic nature of the model manages the complexity of adapting to new patterns

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS20220318681A1System and method for scalable, interactive, collaborative topic identification and tracking

Publication Date: 2022.10.06 CAPITAL ONE SERVICES LLC

AI summary

A topic tracking platform is disclosed that includes a machine-learning model that may be trained to expose topics in a corpus in response to a training table. Because topics are exposed, rather than searched for using existing taxonomies, the sensitivity of a topic tracking platform may be increased, and emerging topic trends may be more quickly flagged. Exposed topics may be automatically labelled, increasing the specificity of the topic tracking platform by overcoming the potential for topic labelling inconsistencies currently experienced in the art. Documents may be scored for each topic using information provided at a token granularity, and the contribution that each token of each document contributes to the topic may be visually represented. In some aspects, mechanisms are provided for reviewing topics of the corpus at varying granularities, including at a topic level, document level or token level granularity.