Data analysis and call prediction on dyadic data from an understudied population

Mehwish Nasim, Aimal Rextin, Shamaila Hayat, Numair Khan, Muhammad Muddassir Malik

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)


In this paper we predict outgoing mobile phone calls using machine learning and time clusters based approaches. We analyze to which extent the calling activity of mobile phone users is predictable. The premise is that mobile phone users exhibit temporal regularity in their interactions with majority of their contacts. In the sociological context, most social interactions have fairly reliable temporal regularity. If we quantify the extension of this behavior to interactions on mobile phones we expect that pairwise interaction is not merely a result of randomness, rather it exhibits a temporal pattern. To this end, we not only tested our approach on an original mobile phone usage dataset from a developing country, Pakistan, but we also analyzed the famous Reality Mining Dataset and the Nokia Dataset (from a European country), where we found an equitable basis for comparison with our data. Our original data consists of 783 users and more than 12,000 active dyads. Our results show that temporal information about pairwise user interactions can predict future calls with reasonable accuracy.

Original languageEnglish
Pages (from-to)166-178
Number of pages13
JournalPervasive and Mobile Computing
Publication statusPublished - Oct 2017
Externally publishedYes


  • Call prediction
  • Call-logs
  • Smartphone
  • Temporal regularity


Dive into the research topics of 'Data analysis and call prediction on dyadic data from an understudied population'. Together they form a unique fingerprint.

Cite this