Democratizing interaction mining

Arsan, Deniz

Permalink

https://hdl.handle.net/2142/122037 Copy

Description

Title

Democratizing interaction mining

Author(s)

Arsan, Deniz

Issue Date

2023-11-30

Director of Research (if dissertation) or Advisor (if thesis)

Kumar, Ranjitha

Doctoral Committee Chair(s)

Kumar, Ranjitha

Committee Member(s)

• Marinov, Darko
• Jabbarvand, Reyhaneh
• Nichols, Jeffrey

Department of Study

Computer Science

Discipline

Computer Science

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Interaction Mining
• Mobile Apps, On-device

Abstract

In the digital landscape, defined by a multitude of mobile applications spanning various platforms, our daily lives are shaped by the quality of digital experiences. The impact of these experiences extends beyond individual satisfaction and directly influences the success of organizations, driving the need for data-driven methods to evaluate and enhance user interfaces. Traditional approaches like analytics and A/B testing, while valuable, require access to an application’s codebase, limiting their utility for external applications. In response to these limitations, interaction mining has emerged as a potent technique. Interaction mining entails the capture of design and interaction data as users engage with an application, resulting in the creation of interaction traces. However, existing interaction mining systems rely on intricate OS-level interventions to enable comprehensive data capture. This dissertation introduces On-Device Interaction Mining (ODIM), a framework that democratizes interaction mining. odim enables data capture by enabling in-the-wild interaction data collection from any Android app on personal devices, without the need for specialized hardware or modifications to the operating system. ODIM empowers researchers, designers, and industry practitioners to enhance task automation, ensure robust user privacy, and bridge the gap between analytics and UX testing. These contributions provide the tools and methodologies needed to innovate digital experiences while safeguarding user privacy. ODIM is a step towards a more accessible, principled, and privacy-conscious future for interaction mining.

Graduation Semester

2023-12

Type of Resource

Thesis

Copyright and License Information

Copyright 2023 Deniz Arsan

Owning Collections