Generating highlights automatically from text-reading behaviors on mobile devices

Abstract

The amount of accessible information online is rapidly expanding. This increases the demand for a support tool that can help people acquire information in a short period of time. Highlights can help with reading, especially by enhancing the effectiveness of skim reading. This paper suggests a novel method of collecting reading-pattern data from a group of subjects through the use of mobile devices so that this data can be utilized to automatically generate highlighted text. Through experimentation, we can gather data while participants read on their mobile devices. The information obtained includes the location of sentences on the screen, the position of scrolling fingers, and the exposure time of each sentence on the screen. From suggested these data, we will develop methods for collecting and filtering the reading patterns of the test subjects, we expect to develop the idea and uses this information for automatically generating the highlights.