App Store Ranking via User Graph and Trust Scores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The discoverability of applications in application stores is hindered due to the lack of inherent linkage among digital content, making it difficult to apply ranking techniques like PageRank, which relies on links between nodes in a linked database, to effectively recommend relevant applications to users.
Innovation Solution
A system that constructs a graph connecting applications through their users, using reputation and trust scores based on social networking data to generate search ranking scores, thereby establishing a connection between applications and recommending relevant content to users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If word matching technique is used to return applications based on search terms, then search relevance is improved, but application discoverability deteriorates due to large number of applications in the store
Solution Approach 1:
Users are introduced as intermediary entities that connect applications through their installation relationships. By computing reputation scores and trust scores for users, the system creates indirect connections between applications, enabling PageRank-like ranking without direct application-to-application links.
Solution Approach 2:
The system transitions from a one-dimensional application database to a multi-dimensional graph structure by incorporating user dimensions. Applications are connected through users who installed them, creating a new dimensional space for ranking that combines application metadata with user reputation and trust metrics.
2Measurement precision
If PageRank algorithm is applied to rank applications, then application importance ranking is improved, but the lack of inherent linkage among digital content deteriorates the ability to apply the algorithm
Solution Approach 1:
Users serve as mediators that provide the necessary linkage structure for PageRank. Instead of requiring direct application links, the system uses user installation records to create indirect connections, allowing PageRank to function in the application store context.
Solution Approach 2:
The system replaces the mechanical link structure required by traditional PageRank with a social graph structure based on user relationships and installation patterns. Reputation scores and trust scores substitute for traditional link weights, enabling the algorithm to operate without direct content linkages.
3Measurement precision
If reputation scores and trust scores are computed for users to generate rankings, then application recommendation accuracy is improved, but computational complexity deteriorates
Solution Approach 1:
Reputation scores and trust scores are computed and stored in advance for users based on their social networking profiles and installation patterns. This preliminary computation allows the ranking system to reuse these pre-calculated values during search operations, reducing real-time computational complexity.
Solution Approach 2:
The system computes scores for a subset of users who have installed candidate applications rather than all users in the system. This partial computation approach reduces the overall computational burden while still providing sufficient data for accurate ranking.
Data Source
AI summary
Disclosed is a technique for providing a linkage between applications, or other digital content, of an application store by assigning values to users of the digital content with respect to a particular user who is submitting a query for particular digital content. The technique involves receiving a query and identifying a subset of digital content based on the query and a word matching technique. For each identified digital content, users who have installed or downloaded the digital content may be determined. A reputation score, corresponding to a user's influence, and a trust score, corresponding to a degree of separation between the user who submitted the query and other users who have installed the digital content, may be computed for each of the users. A list of recommended applications may be generated based on the word matching technique, the reputation score, and the trust score.