Web browsing behavior analysis and interactive hypervideo

Abstract

Processing user interaction data is well-known to be cumbersome and mostly time-consuming, specially when it comes to web browsing behavior analysis. Current tools usually display user interactions as mouse cursor tracks, a video-like visualization scheme that allows researchers to easily inspect what is going on behind the gathered data. However, to date, traditional online video inspection has not explored the full capabilities of hypermedia and interactive techniques. In response to this need, we have developed SMT2ǫ, a web-based tracking system to analyze browsing behavior using feature-rich hypervideo visualizations. We compare our system to related work in academia and industry, showing that ours features unprecedented visualization capabilities. We also show that SMT2ǫ efficiently captures browsing data, and is perceived to be helpful and easy to use. A series of prediction experiments illustrate that raw cursor data are both accessible and easily manipulable, providing evidence that the data can be used to construct and verify hypotheses. Considering its limitations, it is our hope that SMT2ǫ will assist researchers, usability practitioners, and other professionals interested in understanding how users browse the Web.

The Attentive Cursor Dataset

“…Using this final selection of search queries, we produced the static version of the corresponding Google SERPs and injected custom JavaScript code that allowed us to capture all client-side user interactions. For this, we used E v T rack 3 , an open source JavaScript event tracking library derived from the smt2ϵ mouse tracking system (Leiva and Vivó, 2013 ). E v T rack can capture browser events either via event listeners (the event is captured as soon as it is fired) or via event polling (the event is captured at fixed-time intervals).…”

Learning Efficient Representations of Mouse Movements to Predict User Attention

“…However, this suggests that we can expect a visual representation to perform well, since most of the user interactions would be adequately represented in a fixed-size image. Therefore, we created five visual encodings (Table 1) The mouse cursor data were rendered according to each visual encoding using the Simple Mouse Tracking system [48], which was operated via PhantomJS, 5 a scriptable headless browser. Each mouse cursor trajectory was normalized according to the user's viewport.Finally, no data augmentation or transformation techniques were applied, as often performed in computer vision tasks, and the images were saved as 1280x900 px PNG files.…”

Automatic Transformation of a Video Using Multimodal Information for an Engaging Exploration Experience

“…While the above-mentioned approaches create hypervideo using video content, Leiva and Vivó [43] took a different approach. In their study authors use web page interaction logs with a web page to synthesize an interactive hypervideo to allow a user to visualize webpage usage.…”