Context-Aware Search Query Generation via Text Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current internet search methods are inefficient, particularly on small form factor devices, as users must leave their browsing interface to access a search engine, leading to tedious interactions and lack of specificity in search queries due to the need to manually copy or compose search terms from web pages.
Innovation Solution
A system that allows users to select text on a web page to generate and rank multiple search queries based on the context, using a combination of language models and AI systems to provide relevant candidate queries, enabling seamless integration of browsing and searching without leaving the browsing interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users manually copy and paste text to generate search queries, then search queries can be generated, but the process becomes tedious and inefficient particularly on small form factor devices
Solution Approach 1:
The patent combines the browsing interface and search interface into a single unified interface. When users select text on a web page, the system generates candidate search queries directly within the browsing interface, eliminating the need to switch between applications or manually copy-paste text. This merging of functions resolves the contradiction by maintaining efficiency while simplifying user interaction.
Solution Approach 2:
The system performs preliminary actions by automatically generating multiple candidate search queries based on the selected text before the user needs to search. The language model pre-processes the selected text and generates ranked candidate queries, so when the user needs to search, the work of query formulation is already done, improving efficiency without increasing operational complexity.
2Productivity
If users leave the browsing interface to access a search engine, then search can be performed, but the alternation between interfaces is inefficient and tedious
Solution Approach 1:
The patent merges the search functionality directly into the browsing interface by displaying candidate search queries within the same interface where users are reading web pages. This eliminates the need to alternate between browsing and search interfaces, resolving the contradiction by enabling fast search execution while preventing time loss from interface switching.
Solution Approach 2:
The system introduces an intermediary element - a set of candidate search queries displayed within the browsing interface - that mediates between the user's reading activity and the search action. This intermediary allows users to quickly select and execute searches without leaving the browsing context, addressing both the need for fast search execution and the need to minimize interface switching time.
3Measurement precision
If users manually compose search queries, then search can be initiated, but the queries lack specificity and have ambiguous relationships between search terms
Solution Approach 1:
The system enables self-service by using a language model to automatically generate multiple candidate search queries with proper specificity and term relationships based on the selected text. The language model analyzes the context and generates queries with appropriate modifiers and relationships without requiring the user to manually formulate them, thus achieving high query specificity while keeping the interface simple.
Solution Approach 2:
The system changes the parameter of query formulation from manual text composition to automated language model generation. The language model transforms the selected text into multiple candidate queries with optimized specificity and term relationships, resolving the contradiction by achieving precise queries through automated processing rather than manual composition.
Data Source
AI summary
Searching is assisted by recognizing a selection of text from a document as an indication that a user wishes to initiate a search based on the selected text. The user is provided with query suggestions based on the selected text and the query suggestions are ranked based on a context provided by the document. The user may select the text by using a mouse, drawing a circle around the text on a touch screen, or by other input techniques. The query suggestions may be based on query reformulation or query expansion techniques applied to the selected text. Context provided by the document is used by a language model and/or an artificial intelligence system to rank the query suggestions in predicted order of relevance based on the selected text and the context.