Research pick: Where’s the app for that? - "Leveraging app features to improve mobile app retrieval"

There’s an app for that…but which one to choose?

The growth of software – colloquially known as apps, meaning applications – for mobile devices such as smart phones and tablet computers has been enormous. Well-known apps are easy to find or users learn of them through word-of-mouth. However, searching for a previously unknown app that perfectly fits one’s needs is not always straightforward.

Now, writing in the International Journal of Intelligent Information and Database Systems, a team from Algeria and France have developed a new approach to searching for apps that homes in on the functionality the user needs by mining not only the app’s description but also the reviews left by users. The team’s approach then scores the results offering the user the most relevant app to match their needs. The team describes their proof of principle as effective and able to perform better than the state-of-the-art retrieval models for app retrieval.

Messaoud Chaa of the University of Bejaia and the Research Center on Scientific and Technical Information, CERIST, colleague CERIST colleague Omar Nouali, Algeria and Patrice Bellot of Aix Marseille University, France, explain that there were around 30 billion app downloads in 2019 and this number is growing with growing smartphone and tablet adoption around the world. In the Google Play Store alone there are almost 3 million apps, while the Apple App Store carries more than 2 million. “An efficient app search system is essential”, the team writes and at the present time, there is no perfect tool for searching for the app you need that you don’t know exists.

The team’s approach using natural language processing (NLP) allows them to obtain a score for each app and its functions that can be searched by the prospective user and matched more precisely to their needs than a simple app name search might offer.

Chaa, M., Nouali, O. and Bellot, P. (2021) ‘Leveraging app features to improve mobile app retrieval’, Int. J. Intelligent Information and Database Systems, Vol. 14, No. 2, pp.177–197.