Detecting and Characterising Mobile App Metamorphosis in Google Play Store

TL;DR

This study introduces app metamorphosis, a phenomenon where mobile apps drastically shift use cases or market positioning over time. It proposes a multi-modal similarity search combining icon, description, and name/developer embeddings to match 2018 apps to 2023 counterparts without relying on app IDs, and validates the method on two Google Play snapshots five years apart. A taxonomy of metamorphosis categories (re-born, re-branded, re-purposed, genre/ratings/revenue changes, transfers, and progressive versions) is established, with a quantitative success score (SS) that reveals re-branding as particularly advantageous, while also uncovering privacy and security risks such as counterfeits. The findings highlight the need for store operators and users to monitor metamorphosis-related changes, given the potential for misleading identities and evolving permission practices. Limitations include analysis of only two snapshots and reliance on manual validation for certain categories, pointing to avenues for broader temporal studies and automated verification.

Abstract

App markets have evolved into highly competitive and dynamic environments for developers. While the traditional app life cycle involves incremental updates for feature enhancements and issue resolution, some apps deviate from this norm by undergoing significant transformations in their use cases or market positioning. We define this previously unstudied phenomenon as 'app metamorphosis'. In this paper, we propose a novel and efficient multi-modal search methodology to identify apps undergoing metamorphosis and apply it to analyse two snapshots of the Google Play Store taken five years apart. Our methodology uncovers various metamorphosis scenarios, including re-births, re-branding, re-purposing, and others, enabling comprehensive characterisation. Although these transformations may register as successful for app developers based on our defined success score metric (e.g., re-branded apps performing approximately 11.3% better than an average top app), we shed light on the concealed security and privacy risks that lurk within, potentially impacting even tech-savvy end-users.

Figures (14)

• Figure 1: Identified app metamorphosis categories and a notable example pair for each category. (*: 2018 app)
• Figure 2: Creation of validation and test sets from 2018 and 2023 datasets.
• Figure 3: Overall methodology for query, key, and value based best match (if it exists - as emphasised in the purple colour path) retrieval for counterpart app identification to an app ID query from 2018 dataset to be matched with indexed 2023 dataset. We utilise Faisis library for creating the indexes archive.
• Figure 4: Majority voting
• Figure 5: Identifiable mappings and regions of interests that are obtainable from our similarity matching algorithm. Area under each pie segment is indicative of the proportion of apps belonging to each category as described in the legend. Note: top 10k, 20k and 50k dashed lines are not to the scale, but all pie charts and the proportions they represent between top10k, 20k and 50k are true-to-the scale for better visualisation.