Personalized Query Auto-Completion Through Purchase-History Reranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing query auto-completion systems on e-commerce sites lack personalization based on user purchase history, failing to optimize search query suggestions effectively.
Innovation Solution
A mechanism that reranks search query candidates using a natural language processing model trained on user purchase history, incorporating context from the user's current search session, to provide personalized suggestions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a sorting model is trained using only prefix input and search query selection data, then the model can provide personalized query suggestions based on search history, but it fails to incorporate purchase behavior patterns for enhanced personalization
Solution Approach 1:
The patent merges two separate data sources: search query data and purchase history data. The training dataset is constructed by combining prefix information, search query data, and purchase history information into a unified training framework, allowing the model to leverage both search behavior and purchasing patterns for personalized query completion
Solution Approach 2:
The patent performs preliminary action by pre-processing and storing purchase history information in advance. Purchase histories are collected, structured, and integrated into the training data preparation phase before model training, ensuring that purchase behavior patterns are readily available when generating personalized query suggestions
2Measurement precision
If query auto-completion is personalized using search history, then search suggestions become more relevant to user search tendencies, but the system cannot capture purchase-related preferences and behaviors
Solution Approach 1:
The patent adds another dimension to the personalization approach by incorporating purchase history data alongside search query data. This multi-dimensional data integration allows the model to capture both search tendencies and purchasing preferences, creating a more comprehensive user profile for query completion
3Device complexity
If the system uses traditional query auto-completion without purchase data, then the system complexity remains low, but the personalization effectiveness is limited
Solution Approach 1:
Purchase history information is pre-processed and structured in advance during the data preparation phase. This preliminary organization of purchase data allows the model to efficiently access and utilize purchase patterns during inference without adding significant computational complexity during query completion operations
Data Source
AI summary
An information processing apparatus acquires a prefix input on an EC (E-Commerce) site by a user, acquires a context representing a feature relating to a search on the EC site by the user, generates a plurality of search query candidates, based on the prefix, and reranks the plurality of search query candidates, using a result generated by inputting the plurality of search query candidates and the context to a natural language processing model trained using information relating to purchases on the EC site by the user.