Addressing the Recommender System Data Solicitation Problem with Engaging User Interfaces

Quang V Dao, Purdue University

Abstract

The motivation for this work comes from the need to obtain data for autonomous systems that rely heavily on recommender systems for human interaction. Recommender systems are data-driven technologies and depend heavily on large quantities of user data to function effectively, e.g. [1], [2]. However, acquiring that data has proven difficult [3]–[7] because users typically do not want to lend the effort to furnish the data. Part of the reason for this problem stems from the reluctance of users to provide data because, as reported by users, it is cognitively taxing and/or too tedious to do so [8], [9]. With autonomous systems bringing greater demand for user data, in some applications, this also brings an opportunity to solicit data from users. The American driver spends about an hour each day on the road; with self-driving cars, this means there will be a captive audience in the vehicle for at least the duration of the trip [10]. To exploit this, a user interface will need to be designed to coax the user into achieving system goals, like data solicitation. One approach is to design a system to leverage an already present tendency for people to socially interact with technology [11]. In this thesis, I argue that such an approach would involve incorporating interaction concepts that facilitate engagement into the design of recommender system interfaces that will improve the likelihood of obtaining data from users. To support this claim, I synthesize past work on human-computer interaction and recommender systems to derive a framework to guide scientific investigations into interface design concepts that will address the data solicitation problem.In addition, I present the results of a study of how anthropomorphism, as an indicator of engagement [12], may affect the amount of data provided by a user. I begin with a discussion of the problem and then provide a description of the recommender system filtering process to illustrate why user data is important. Then I describe the types of data that will be relevant to recommender system functioning, as the type of data determines how it is obtained. Subsequently, I introduce the construct of engagement and discuss design concepts for interactions that can potentially support it. I conclude with a discussion of future empirical work aimed at testing elements of the human-computer interaction approaches presented herein. The contribution will be the framework stated above and a research agenda stemming from that framework.

Degree

Ph.D.

Advisors

Yu, Purdue University.

Subject Area

Design|Communication|Information science|Social psychology|Transportation

COinS