Location Activity Recommendation Service Using Matrix Factorization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Raw GPS data, which consists of coordinates and timestamps, is difficult to interpret to identify locations, places, and events, necessitating a transformation to convey meaningful information.

Innovation Solution

A location and activity recommendation service constructs matrices to represent relationships between locations and activities by detecting GPS stay points, extracting user comments, and correlating points of interest, then applies collective matrix factorization to infer correlations and provide recommendations based on user queries.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If raw GPS data is used directly, then data collection is simple, but it is difficult to identify locations, places, and events

Engineering Contradiction:

Improvedata collection simplicityVSAvoidinterpretability of location and event information

Core Design Contradiction:

Ease of manufactureVSLoss of information

Solution Approach 1:

The patent introduces an intermediary processing system that transforms raw GPS coordinates into meaningful location and event information. The system uses POI databases, user comments, and matrix factorization algorithms as intermediaries to bridge the gap between raw data and interpretable information, resolving the contradiction between data collection simplicity and information interpretability

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces manual interpretation of GPS data with automated computational methods. Matrix factorization algorithms and data processing systems substitute for human analysis, automatically extracting location and event information from raw coordinates, thereby maintaining data collection simplicity while improving information interpretability

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Loss of information

If GPS data is transformed to convey location and event information, then information interpretability improves, but processing complexity increases

Engineering Contradiction:

Improveinterpretability of location and event informationVSAvoiddata processing complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent segments the complex data processing task into distinct components: GPS data collection, stay point detection, POI matching, user comment extraction, and matrix construction. Each component handles a specific aspect of the transformation, reducing overall processing complexity while maintaining information interpretability

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediate data structures (location-activity matrices, POI databases, stay point representations) that facilitate the transformation process. These intermediaries simplify the complex mapping between raw GPS data and meaningful information, making the processing pipeline more manageable and less complex

Inventive Principle:

Principle #24Intermediary (Mediator)

3Measurement precision

If multiple matrices are constructed to represent location-activity relationships, then recommendation accuracy improves, but computational requirements increase

Engineering Contradiction:

Improverecommendation accuracyVSAvoidcomputational energy consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent performs preliminary data processing and matrix construction during off-peak times or in batch operations. By pre-computing location-activity matrices and relationships before they are needed for recommendations, the system reduces real-time computational energy consumption while maintaining high recommendation accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent employs matrix factorization techniques that transform large, complex matrices into smaller factor matrices with fewer parameters. This dimensionality reduction maintains the essential location-activity relationships and recommendation accuracy while significantly reducing computational energy requirements for processing and storage

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS8719198B2Collaborative location and activity recommendations

Publication Date: 2014.05.06 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Techniques describe constructing a location and activity recommendation model to identify relationships between locations and activities. To construct the model, the process obtains global positioning system (GPS) logs of geographical locations collected over time and identifies stay points representing locations visited by an individual user. The process also identifies points of interest in a region using a database and correlates a relationship between activity to activity by submitting queries to a search engine. The information gathered is used to fill locations and activities in a location-activity matrix. Recommendations may be made for a location and/or activity when given a user query, based on a user's present geographical location, or a prediction of a user's interest.