Game Medium Set Recommendation Using Verified Candidate Replacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing game recommendation systems face accuracy issues with low-quality recommendations for novice players due to the 'cold start' problem and 'exposure bias' from intermediate and advanced players, leading to suboptimal game medium set configurations.
Innovation Solution
A system generates candidate game medium sets by replacing game media in cleared sets with higher priority media, executes game plays to verify, and outputs the cleared sets, using virtual instances to simulate player interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing information recommendation systems are used to recommend game medium sets, then the system can provide recommendations to players, but the recommendation accuracy is low for novice players due to the cold start problem
Solution Approach 1:
The system pre-generates candidate game medium sets by replacing some game media in existing cleared sets with other game media before actual game plays occur. This preliminary generation of candidates based on game logs allows the system to have recommendation candidates ready even when clearing history data is scarce, thus addressing the cold start problem for novice players.
Solution Approach 2:
The system introduces virtual instances as intermediaries to execute game plays and verify candidate game medium sets. These virtual instances act as a mediator between the recommendation system and actual player gameplay, allowing the system to evaluate candidate sets without requiring extensive real player data, thereby improving recommendation accuracy for players with limited clearing history.
2Productivity
If recommendation systems simply recommend based on clearing histories of other users, then the system can provide recommendations quickly, but exposure bias occurs where intermediate and advanced players adopt recommended configurations making only a very small number of party configurations established tactics
Solution Approach 1:
The system dynamically generates multiple candidate game medium sets by replacing different game media in existing cleared sets, rather than simply copying existing configurations. This dynamic generation process ensures that recommendations are adapted and varied, preventing the concentration on a single established tactic while maintaining quick recommendation delivery.
Solution Approach 2:
The system executes game plays using virtual instances to verify whether candidate game medium sets can actually clear the target unit game. This feedback mechanism ensures that only valid and effective configurations are recommended, maintaining recommendation quality and diversity without sacrificing speed, as the verification process is automated and efficient.
3Measurement precision
If the system generates candidate game medium sets by replacing game media and executes game plays to verify, then the accuracy of recommendations is improved, but the computational complexity and processing time increase
Solution Approach 1:
The system segments the recommendation process into distinct modules: generating candidate sets by replacing specific game media, executing game plays using virtual instances, verifying clearance results, and outputting final recommendations. This segmentation allows each module to be optimized independently, managing computational complexity while maintaining high recommendation accuracy through systematic verification.
Data Source
AI summary
One or more embodiments of the invention is a system that outputs a game medium set to be presented to a player in a game in which game medium sets each constituted of a plurality of game media are used, wherein the system: generates candidate game medium sets for clearing one unit game by replacing, with other game media, some game media in game medium sets with which the one unit game has been cleared, the game medium sets being included in game logs; and executes game plays of the one unit game by using the candidate game medium sets for clearing the one unit game and outputs a game medium set with which the one unit game has been cleared.