Abstract
The number of people with access to mobile devices, as well as applications to these devices (i.e., apps), has been increasing significantly. Thus, users have to choose among a large number of apps proposing to do the same functions, those that better serve them. A possible solution to this problem is the adoption of recommendation systems. Meanwhile, usually these systems consider only users' preferences to create a profile or request sensitive data (e.g., call and message logs). This work investigates the impact of using demographic and device information on app recommendation by using only easy-to-obtain data to enrich a user profile. We evaluate two approaches: a similarity-based Collaborative Filtering with a limited number of apps and a topic-based approach (i.e., LDA) with wider large-scale data. We also inspected the results under both apps and categories context. The general results reveal that the enriched data provides a better app recommendation with the addition of information about the user's region mean wage achieving up to 210% (or 12 percentage points) of improvement in terms of recall.
Publisher
Sociedade Brasileira de Computacao - SB
Subject
Computer Networks and Communications,Computer Science Applications
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Extracting Mobile User Profile using Easy-to-obtain and Less Invasive Data;Proceedings of the Int'l ACM Symposium on Performance Evaluation of Wireless Ad Hoc, Sensor, & Ubiquitous Networks;2023-10-30