User History Query Retrieval for Natural Language Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search systems lack the ability to allow users to locate information they have previously accessed using natural language queries.
Innovation Solution
A search system that receives a natural language query, classifies it as seeking previously accessed information, applies filters based on the query, and retrieves relevant documents from the user's history, such as browser history or email, to generate search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional search systems provide search results from a database, then users can access general information, but users cannot efficiently locate information they have previously accessed using natural language queries
Solution Approach 1:
The system performs preliminary actions by storing and indexing user access history (documents, URLs, emails, attachments) before queries are submitted. This pre-processing enables rapid retrieval when users submit natural language queries about previously accessed information, eliminating the need to search through entire databases during query execution.
Solution Approach 2:
The patent introduces an intermediary classification layer that sits between the natural language query interface and the search database. This intermediary component classifies queries to determine if they reference previously accessed information, then routes them to the appropriate user history database, enabling the system to handle natural language queries without requiring complex parsing of user intent.
2Adaptability or versatility
If the system stores and processes user access history, then personalized search results can be generated, but system complexity increases
Solution Approach 1:
The system segments user history into distinct categories (documents accessed, URLs visited, emails read, attachments viewed) with separate storage structures for each type. This segmentation allows the system to handle different data types with appropriate specialized processors, reducing overall system complexity while enabling comprehensive personalization across multiple information sources.
Solution Approach 2:
The patent implements a universal query processing framework that handles multiple types of user history data (documents, URLs, emails, attachments) through a single integrated system. This multi-functional approach allows the same infrastructure to serve multiple purposes - storing diverse data types, classifying various query formats, and retrieving different information kinds - thereby reducing complexity compared to maintaining separate systems for each function.
3Measurement precision
If the system classifies natural language queries to identify user history references, then accurate retrieval of previously accessed information is enabled, but processing time increases
Solution Approach 1:
The system applies partial classification action by using the classification layer to determine only whether a query references previously accessed information, without performing complete semantic analysis of the entire query. This partial processing approach achieves sufficient accuracy for routing decisions while minimizing processing time, avoiding the need for exhaustive natural language understanding for every query.
Solution Approach 2:
The system performs preliminary classification of queries to quickly determine if they reference user history before initiating full search operations. This preliminary action filters out queries that don't require history search, reducing the number of queries that undergo time-consuming full processing while maintaining accurate classification for those that do require history retrieval.
Data Source
AI summary
One or more servers receive a natural language query from a client device associated with a user. The one or more servers classify the natural language query as a query that seeks information previously accessed by the user. The one or more servers then obtain a response to the natural language query from one or more collections of documents, wherein each document in the one or more collections of documents was previously accessed by the user. The one or more servers generate search results based on the response. Then, the one or more servers communicate the search results to the client device.