Crowd Data Filtering for Point-of-Interest Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for displaying Points-of-Interest (POIs) lack the ability to consider user surroundings and affinity, failing to account for newly established locations and not providing adequate filtering based on crowd data or surroundings.
Innovation Solution
A system and method that filters POIs using crowd data, identifying and removing POIs with unsatisfactory crowd attributes, and creates new POIs based on crowd-sourced requests, utilizing a Mobile Aggregate Profile (MAP) system that aggregates user profiles and crowd data to provide filtered and relevant POI lists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If POIs are displayed statically or based on simple user input, then the system is simple and easy to operate, but the system cannot consider user surroundings and affinity, resulting in lower relevance
Solution Approach 1:
The system pre-establishes a POI database with comprehensive attributes and categorizes crowds into different types (e.g., locals, tourists, commuters) with predefined characteristics. This preliminary preparation enables the system to quickly match users with relevant POIs without complex real-time analysis, thereby improving recommendation relevance while maintaining operational simplicity.
Solution Approach 2:
The patent introduces 'crowd data' as an intermediary element that bridges user preferences and POI attributes. By analyzing crowd characteristics at potential POIs and comparing them with user profiles, the system indirectly determines POI relevance without requiring direct complex user-POI matching, thus enhancing measurement precision while managing system complexity.
2Adaptability or versatility
If the system uses predefined categories for filtering POI, then the operation is simple, but the system cannot identify newly established POIs or POIs with desirable surroundings
Solution Approach 1:
The system implements multiple filtering stages: first applying simple predefined category filters for quick results, then optionally applying additional crowd-based filtering for users seeking more specific matches. This partial application of complex filtering only when needed maintains ease of operation for simple queries while enabling comprehensive POI discovery when users require adaptability.
Solution Approach 2:
The system pre-collects and stores crowd data for POIs including attributes like crowd type, density, and characteristics. This preliminary data gathering enables the system to automatically identify newly established POIs and those with desirable surroundings without requiring complex real-time analysis, thus improving adaptability while keeping the user interface simple.
3Measurement precision
If the system filters POIs based on crowd data and attributes, then the recommendation accuracy improves, but the processing time and computational resources increase
Solution Approach 1:
The filtering process is divided into segments: first filtering POIs by basic categories and user preferences, then applying crowd-based filtering only to the reduced subset. This segmentation reduces the computational burden of crowd data analysis while maintaining high matching accuracy, as crowd filtering is applied only where needed rather than to all POIs universally.
Solution Approach 2:
The system applies comprehensive crowd-based filtering selectively based on user needs and query context. For simple queries, basic filtering suffices and saves time. For complex queries where accuracy is prioritized, the full crowd-based filtering is applied. This partial application optimizes the balance between processing time and matching accuracy.
Data Source
AI summary
Systems and methods are provided for filtering and/or creating Points-of-Interest (POIs). In one embodiment, a list of POIs is obtained and then filtered based on crowd data related to the list of POIs to provide a filtered list of POIs. In another embodiment, one or more crowd-sourced POIs are created based on one or more crowds relevant to a corresponding bounding region for POI creation.