The Reader's Helper: A Personalized Document Reading Environment

Document Annotation.

Figure 2 shows the HTML document from figure 1 after it has been annotated by the annotation agent. The locations of relevant phrases in the document are shown in red in the Thumbar. Also, the pattern of bolded words in figure 2 matches the highlighting pattern in the Thumbar lens. Using the lens the reader can quickly reposition the document to areas containing relevant phrases. Notice also that there are four concepts with non-zero scores in this document. Each concept has a score in the label and a grid meter that is populated according to the value of the score. By simply looking at the collection of concepts the reader can quickly see that this document is indeed similar to several topics of interest based on which grid meters have color and to what extent. This is a important but subtle part of the interface design. This is based on the belief that readers will quickly learn the location of concepts in the concept area and as a consequence of highlighting in the grid meter, quickly know if the document discusses the desired topics of interest and which topics are covered. Note that each concept is also a button which can be turned on or off. For instance, the user may choose to only view the annotations for Wearable, turning off all other scoring concepts. Concepts may also be turned off prior to analyzing a document.

Highlighting Styles

Three highlighting styles are provided for use while reading a document . These styles currently include 1) highlighting only the phrase, 2) bolding the phrase and highlighting the sentence in which it appears, and 3) underlining the phrase. The reader may select the highlighting style at anytime, and personalize the color schemes and styles in the reader profile.

Sensitivity Meter

Also represented in figure 2 is the sensitivity threshold meter . This meter can be used to set the threshold for the document area and Thumbar, restricting the annotation to concepts scoring above the specified threshold. For instance, the concept Design currently has a score of 76%. The reader could use the sensitivity meter to increase the threshold to a value greater than 76% and all annotations associated with the Design concept would disappear from the display. This gives the user control of showing only the higher scoring concepts when reading a document, further supporting the concept of personalizing the document.