Natural Language Content System with Corrective Feedback

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing content search and discovery systems face challenges in accurately interpreting user natural language queries due to limited user data and indirect relationships between user intent and search results, leading to inefficient use of network resources and decreased user satisfaction.

Innovation Solution

A natural language-based content system with corrective feedback and training service that generates training samples based on user feedback to update natural language understanding logic, using a distribution function to select query objects and a multi-interpretative framework for statistical and linguistic analysis to improve search relevance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional search and recommendation systems are used to filter and present content titles, then user satisfaction depends on result relevance, but the time and effort required to formulate queries increases

Engineering Contradiction:

Improvesearch result relevanceVSAvoidquery formulation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system implements feedback loops where user interactions with search results (clicks, views, selections) are continuously collected and used to refine the natural language understanding model. This feedback mechanism allows the system to learn from user behavior patterns and improve query interpretation over time, reducing the need for users to spend time formulating precise queries while maintaining high result relevance

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The natural language understanding system automatically interprets and processes user queries without requiring users to manually refine or rephrase their search terms. The system self-adjusts by learning from interaction data, enabling it to accurately understand intent from initial queries and present relevant results, thereby minimizing the time users spend formulating queries

Inventive Principle:

Principle #25Self-service

2Extent of automation

If limited user data is used for statistical-based learning, then natural language understanding can be implemented, but the accuracy of tag generation and search results decreases

Engineering Contradiction:

Improvenatural language understandingVSAvoidsearch result accuracy

Core Design Contradiction:

Extent of automationVSMeasurement precision

Solution Approach 1:

The system performs preliminary actions by collecting and preprocessing user interaction data continuously in the background before actual search queries are processed. This ongoing data collection and model training occurs proactively, allowing the system to build up sufficient training data and improve its natural language understanding capabilities ahead of time, thereby maintaining high search accuracy even with initially limited data

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system maintains continuous learning and improvement through ongoing collection of user interaction data and iterative model training. Rather than relying on a static dataset, the system continuously refines its natural language understanding and tag generation accuracy by processing new interaction data, ensuring sustained high performance in search result accuracy while maintaining full automation

Inventive Principle:

Principle #20Continuity of useful action

3Adaptability or versatility

If comprehensive search results are generated for user queries, then content discovery options increase, but network resource usage increases

Engineering Contradiction:

Improvecontent discovery optionsVSAvoidnetwork resource usage

Core Design Contradiction:

Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The system applies local quality by generating search results with varying levels of detail and comprehensiveness based on specific query characteristics, user preferences, and contextual factors. Rather than uniformly generating comprehensive results for all queries, the system tailors the extent of search processing to each specific case, providing detailed results when needed while using fewer resources for simpler queries, thereby maintaining content discovery options while optimizing network resource usage

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system dynamically adjusts search parameters such as result depth, filtering criteria, and processing intensity based on query analysis and learned user behavior patterns. By changing these parameters adaptively, the system can provide comprehensive content discovery options when appropriate while reducing network resource consumption for queries that don't require extensive processing, effectively balancing versatility with resource efficiency

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11636102B2Natural language-based content system with corrective feedback and training

Publication Date: 2023.04.25 VERIZON PATENT & LICENSING INC

AI summary

A method, a device, and a non-transitory storage medium are described, which provide a natural language-based content system with corrective feedback and training service. The natural language-based content system with corrective feedback and training service may collect data based on interaction with search results from users. The natural language understanding model may generate feedback data based on the collected data, and use the feedback data to further train the natural language understanding model and update search and discovery logic for searching and discovering contents. The feedback data may categorize errors based on the interaction, and identify differences between search queries received during a search session with a user.