Personalized Search Suggestions via Key Phrase Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines fail to anticipate user needs by not providing personalized search suggestions based on past content viewed, limiting users to only their past search history and not offering content similar to previously viewed material.
Innovation Solution
A system that generates personalized search suggestions without user input by extracting key phrases from previously viewed website content, assigning confidence scores, and pre-populating search queries in the user interface, using natural language processing and machine learning to determine user interests and suggest relevant content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If search engines provide only past search history for selection, then users can access previously searched content, but users cannot access similar content they have not yet searched for
Solution Approach 1:
The system performs preliminary analysis of user browsing behavior by extracting key phrases from previously viewed web pages and pre-processing this information into structured data. This preliminary action enables the system to generate personalized search suggestions before the user actually performs a new search, anticipating user needs rather than merely responding to past searches.
Solution Approach 2:
The system introduces an intermediary processing layer between raw browsing history and search suggestions. This intermediary component extracts key phrases from web page content, analyzes user browsing patterns, and transforms unstructured browsing data into structured search suggestions. This intermediary process enables the system to infer user interests and generate relevant content recommendations without direct user input.
2Productivity
If search engines require users to enter new search queries, then users can specify precise search needs, but users must repeatedly enter the same search terms
Solution Approach 1:
The system enables self-service by automatically generating search suggestions based on user browsing behavior without requiring explicit user input. The system monitors user interactions with web pages, extracts relevant key phrases, and autonomously creates personalized search suggestions that reflect user interests. This self-service mechanism eliminates the need for users to manually re-enter search terms while maintaining search precision.
Solution Approach 2:
The system implements feedback loops by continuously monitoring user interactions with search suggestions and adjusting future recommendations based on user preferences. When users interact with suggested searches or browse certain types of content, the system uses this feedback to refine key phrase extraction and improve the relevance of future search suggestions, creating a dynamic adaptation to user needs.
3Adaptability or versatility
If search engines provide generic search suggestions, then suggestions can be provided for any user, but suggestions are not personalized to individual user interests
Solution Approach 1:
The system segments user browsing behavior into discrete key phrases extracted from specific web pages. By breaking down complex browsing histories into individual key phrases and associated metadata, the system can analyze and process user interests in manageable units. This segmentation enables personalized suggestion generation by identifying patterns in specific key phrases rather than attempting to analyze entire browsing histories as monolithic data structures.
Solution Approach 2:
The system changes parameters by extracting and weighting different key phrases based on user browsing patterns. Instead of treating all search suggestions equally, the system adjusts parameters such as key phrase frequency, recency, and relevance to generate personalized suggestions. This parameter-based approach allows the system to adapt suggestion generation to individual users while maintaining a relatively simple underlying architecture.
Data Source
AI summary
Examples of the present disclosure describe systems and methods that provide a pipeline to generate personalized queries that are associated with and based on a user's interests determined from a user's past searches, on an Internet search engine, and/or the content the user viewed from the past searches. The suggested queries can be shown in a user interface component associated with the user interface of the search engine and before the user enters anything, such as a new Internet search. This pre-population of searches associated with a user's interests gives an opportunity to the user to try these queries without manually entering in a search string.