Time-Series User-Item Graphs for Accurate Behavior Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing knowledge graphs do not effectively incorporate time-series relationships between users and items, which affects the accuracy of customer behavior analysis.
Innovation Solution
A knowledge graph construction method that includes a feature acquisition unit, a time-series information acquisition unit, and an extraction unit to create a graph with user and item nodes and links that reflect time-series interactions, allowing for the extraction of node representations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a knowledge graph includes both users and items as entities with static relationships, then the structure is simple and easy to construct, but the accuracy of user behavior analysis is insufficient because time-series relationships are not reflected
Solution Approach 1:
The patent transforms the static knowledge graph into a dynamic structure by incorporating time-series information. The construction unit creates a knowledge graph where relationships between users and items are weighted based on temporal patterns, allowing the graph to reflect changing interaction strengths over time. This dynamic approach enables the system to capture that recent interactions are more significant than historical ones, thereby improving analysis accuracy while maintaining manageable complexity through systematic time-based weighting.
2Reliability
If time-series information is incorporated into the knowledge graph, then the accuracy of user behavior analysis is improved, but the construction process becomes more complex
Solution Approach 1:
The patent applies preliminary action by pre-processing time-series interaction data before graph construction. The construction unit systematically processes historical interaction records, calculates time-based weights for each relationship, and incorporates these weights into the knowledge graph structure. This pre-computation of temporal weights simplifies the overall construction process by breaking down the complex task into manageable steps: data collection, time-weight calculation, and graph assembly, thereby maintaining reliability while improving ease of manufacture.
Data Source
AI summary
An information processing apparatus acquires user features indicating features for each of a plurality of users and item features indicating features for each of a plurality of items, acquires time-series information indicating time-series interactions between the plurality of users and the plurality of items, constructs, based on the user features, the item features, and the time-series information, a graph including a plurality of user nodes representing the plurality of users, a plurality of item nodes representing the plurality of items, and links indicating interaction relationships in the plurality of user nodes and the plurality of item nodes, and extracts, from the graph, a node representation in the graph for any node among the plurality of user nodes and the plurality of item nodes.