EyePrint: Using Passive Eye Trace From Reading to Enhance Document Access and Comprehension

Abstract

The improvement of digital document technologies is changing our document browsing style. We browse a lot of documents in a short time, jumping from one to another by following the hyperlinks. This makes document management difficult; we sometimes feel it is difficult to keep track of all of the information. To overcome this problem, this article proposes a novel method that uses the user's eye movement to create passive traces of document comprehension for digital documents. The eye gaze traces so generated become metadata of the document; they can be used in the next search, document browsing, and zooming. A prototype system that works with a gaze tracking system is developed. Two experiments are conducted to evaluate the usefulness of eye gaze traces in digital document browsing. The first experiment examines the validity of trace generation, where generated eye gaze traces are compared with the ‘participants’ self-judgment of reading. In the second experiment, participants browse a set of documents to gather information, where the ‘participants’ reading behavior with the eye gaze trace and the efficiency of keyword search is studied. The results confirm that EyePrint is the answer to keeping track of digital documents.

Acknowledgments

I thank Dr. Hisao Nojima, professor at Seijo University, for his contribution to the original idea of this research. This project came from a discussion with him. I am also grateful to the members of the Human Appliances Project of Cyber Solutions Laboratories for their invaluable comments on this research.

Notes

¹Japanese morphological analysis system ChaSen, http://chasen.aist-nara.ac.jp/

²In the first stage of Experiment 2, participants visited 85% pages on average. Therefore, the number of search results was reduced by approximately 15%, which contributes to improve the precision by 18% (1/0.85 = 1.18).