Application Recommendation Server Using Terminal Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing application recommending methods often result in applications being incompatible with terminals when run, due to the diverse range of operating systems and hardware configurations.
Innovation Solution
An application recommending method and system that involves receiving data from terminals, calculating activity indices of applications based on their behavior data, and recommending applications with high activity indices to ensure compatibility and smooth operation on the terminal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If application recommending is performed across all terminal types, then application coverage is improved, but compatibility reliability deteriorates
Solution Approach 1:
The patent segments terminals into different types based on static data characteristics (OS version, hardware configuration, manufacturer, model). By dividing the heterogeneous terminal population into homogeneous groups, the system can recommend applications within each segment where compatibility is more predictable, thus maintaining reliability while achieving broad coverage across all segments.
Solution Approach 2:
The patent applies local quality by calculating activity indices separately for each terminal type rather than using a universal index. This allows the recommendation system to adapt the evaluation criteria to local characteristics of each terminal segment, improving compatibility within each group while maintaining overall coverage across diverse terminals.
2Measurement precision
If activity index calculation includes comprehensive behavior data, then recommendation accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent extracts only the essential behavior data elements needed for activity index calculation (installation time, start time, activation time, deactivation time, stop time, uninstallation time) from the comprehensive terminal data. By selecting and extracting only the relevant temporal metrics, the system achieves accurate recommendation without processing unnecessary complex data.
Solution Approach 2:
The patent implements a dynamic activity index calculation that adapts to different terminal types and application categories. The system dynamically adjusts which behavior data elements are weighted more heavily based on terminal characteristics and application types, achieving high accuracy without fixed complex processing rules.
Data Source
AI summary
Embodiments of the present disclosure are applicable to the field of communications technologies, and provide an application recommending method and system, and a server. The method includes receiving data reported by at least one terminal, where the data includes static data and first application behavior data that are collected by the at least one terminal, and the static data is used to identify a type of a terminal; obtaining an activity index of each application on terminals of different types according to the first application behavior data; and receiving an application list request sent by a first terminal, querying, according to the application list request, an activity index of each application on terminals that are of the same type with the first terminal, and recommending an application with an activity index greater than a preset first activity index threshold to the first terminal.