Multi-item Influence Maximization via Item Association Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional influence maximization techniques focus on maximizing user coverage for a single product by equally allocating resources to a small subset of social influencers, failing to effectively promote multiple items simultaneously and efficiently allocate resources across a social network.
Innovation Solution
Implementing a multi-item influence maximization system using a mean-field approximation of the Ising model to iteratively determine resource allocation for users based on both social and item influences, allowing for variable resource distribution across users to maximize the spread of influence across multiple items.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional influence maximization techniques are used to maximize user coverage for a single product, then user coverage is improved, but resource allocation efficiency deteriorates when promoting multiple items simultaneously
Solution Approach 1:
The patent segments the resource allocation problem by introducing item association graphs that divide the social network into item-specific influence pathways. Instead of treating all users uniformly, the system segments resource allocation based on item correlations and user-item associations, allowing efficient targeting of resources to specific items while maintaining overall multi-item promotion effectiveness.
Solution Approach 2:
The patent adds a new dimension to the influence maximization problem by incorporating item association graphs alongside traditional social correspondence graphs. This dimensional expansion allows the system to simultaneously optimize for multiple items by considering both social relationships and item correlations, thereby improving resource allocation efficiency without sacrificing user coverage.
2Ease of operation
If equal portions of budget are allocated to determined social influencers, then implementation simplicity is improved, but influence maximization effectiveness deteriorates
Solution Approach 1:
The patent applies local quality by allocating resources differently to different users based on their specific influence characteristics and item associations. Instead of uniform allocation, the system determines variable resource portions for each user-item pair based on local properties such as user influence scores, item popularity, and association strengths, thereby maximizing influence spread while maintaining manageable complexity through automated calculations.
3Device complexity
If discrete budget settings are used for influence maximization, then computational simplicity is improved, but optimization precision deteriorates
Solution Approach 1:
The patent introduces dynamics by transitioning from discrete to continuous resource allocation. The system uses continuous variables to represent resource portions allocated to each user-item pair, allowing for precise optimization through iterative algorithms. This dynamic approach enables fine-grained adjustment of resource distribution while maintaining computational feasibility through efficient gradient-based optimization methods.
Data Source
AI summary
In implementations of multi-item influence maximization, a computing device can obtain updates to a user association graph that indicates social correspondence between users, and obtain updates to a user-item graph that indicates user correspondence with one or more items. The computing device includes an influence maximization module that can update an item association graph that indicates item correspondence of each item with one or more other items, where the item association graph can be updated based on the user-item graph that indicates the user correspondence with one or more of the items. The influence maximization module can then iteratively determine a resource allocation for each of the users to maximize user influence of multiple items that are associated in the item association graph and based on the social correspondence between the users, as well as assign a variable portion of the resource allocation to any number of the users.