Learning users' interests by unobtrusively observing their normal behavior

TLDR

Learning user interests is costly because it requires user labeling, yet prior work has relied on passive observation of clicks, and we assume that easily measurable actions such as clicks, scrolling, and mouse movements correlate strongly with interests. The study proposes a method that eliminates the need for human‑labeled pages by using surrogate tasks based on observable user actions. By unobtrusively recording clicks, scrolling, and mouse activity, the system generates surrogate training data, learns predictive functions, and then uses these functions to recommend likely‑interest pages during idle periods. Empirical results demonstrate that the agent accurately predicts easily measurable aspects of browser usage, validating the surrogate‑task approach.

Abstract

For intelligent interfaces attempting to learn a user's interests, the cost of obtaining labeled training instances is prohibitive because the user must directly label each training instance, and few users are willing to do so. We present an approach that circumvents the need for human-labeled pages. Instead, we learn "surrogate" tasks where the desired output is easily measured, such as the number of hyperlinks clicked on a page or the amount of scrolling performed. Our assumption is that these outputs will highly correlate with the user's interests. In other words, by unobtrusively "observing" the user's behavior we are able to learn functions of value. For example, an intelligent browser could silently observe the user's browsing behavior during the day, then use these training examples to learn such functions and gather, during the middle of the night, pages that are likely to be of interest to the user. Previous work has focused on learning a user profile by passively observing the hyperlinks clicked on and those passed over. We extend this approach by measuring user mouse and scrolling activity in addition to user browsing activity. We present empirical results that demonstrate our agent can accurately predict some easily measured aspects of one's use of his or her browser.