Map Points of Interest Display Using Predicted Popularity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for displaying Points of Interest (POI) on maps are inaccurate due to reliance on expert-selected rules that are limited in applicability, leading to difficulty in identifying user-interest points, especially in diverse application scenarios.
Innovation Solution
A method that acquires features of candidate POIs, determines predicted popularity based on user operation frequency over historical time periods, and displays POIs meeting a preset popularity condition, using a mapping relation between features and popularity to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If expert-selected rules are used to display POIs on maps, then the implementation is simple, but the accuracy of identifying user-interest POIs deteriorates
Solution Approach 1:
The patent replaces expert-based manual selection rules with an automated machine learning model that processes user operation data. The model learns mapping relations between POI features and user interest automatically, substituting human expert judgment with data-driven algorithms to improve accuracy while maintaining implementation feasibility through standardized ML pipelines.
Solution Approach 2:
The system enables POI selection to serve itself by automatically learning from user behavior data without requiring continuous expert intervention. The model self-adjusts by processing historical user operations and automatically updates POI recommendations, making the system self-improving and reducing dependency on external expert knowledge.
2Device complexity
If expert-selected rules are used to display POIs on maps, then the system complexity is low, but the adaptability to diverse application scenarios deteriorates
Solution Approach 1:
The patent creates a universal POI selection system that handles multiple application scenarios through a single machine learning model. The model processes various types of user operation data (searches, navigation, favorites) and adapts to different scenarios (travel, daily commute, shopping) by learning general mapping relations between POI features and user interest, eliminating the need for scenario-specific rule sets.
Solution Approach 2:
The system achieves adaptability by dynamically adjusting model parameters based on learned mapping relations from user data. Instead of fixed expert rules, the model modifies its internal parameters (weights, thresholds) according to observed user behavior patterns, enabling flexible adaptation to diverse scenarios while maintaining a consistent system architecture.
3Use of energy by moving object
If traditional POI display methods are used, then the computational resources required are minimal, but the relevance of displayed POIs to user interest deteriorates
Solution Approach 1:
The patent implements partial processing by selecting only the most relevant POIs for display based on predicted user interest scores. The model processes user operation data to generate interest predictions, then filters and displays only the top-ranked POIs that meet relevance thresholds, avoiding unnecessary computation on irrelevant data while maintaining high information quality.
Data Source
AI summary
The present disclosure discloses a method and apparatus for displaying map points of interest, and an electronic device, relates to the field of artificial intelligence, and in particular to intelligent transportation. A specific implementation solution includes: acquiring features corresponding to multiple candidate points of interest; determining predicted popularity of the multiple candidate points of interest according to a mapping relation between each feature and each popularity and the features of the multiple candidate points of interest, and the mapping relation is determined based on the frequency of operations performed by a user for each sample point of interest in a historical time period; and displaying the candidate points of interest of which predicted popularity meets a preset popularity condition in a map. Therefore, the accuracy of the displayed points of interest may be enhanced.