Have A Say Over What You See: Evaluating Interactive Compression Techniques, IUI 2009

When presented with as many documents and written data as we are today, it becomes almost impossible for us to read everything line by line.  Many people are able to skim the document relatively quickly but in doing this much information could be lost.  Tools for what the paper calls Interactive Compression, IC, could greatly help the reader better identify important parts of the document on which to focus.  This paper highlights some functions and ideas from IC.

The paper attempts to identify effective principals for IC systems by looking at the following attributes; Comparisson with unmodiffied text, efficiency, accuracy, granularity, and interactivity.  Using these principals the authors identified the two best forms of IC, word excision and word highlighting.  Word excision will attempt to remove unimporatnt phrases and words where word highlighting attempts to hightlight the important words to aid a reader in skimming the document.

In a user study the word highlighting method was recieved better than the word excision method.  We can propose that while these methods are similiar it would be fair to say that a user would most likely desire to still have all the information presented to him or her so that they can read in more depth for certain sections.  This, while possible with word excision, is not as easy as with the word highlighting method.

The authors found that while some key elements were missed the systems as a whole were successful.  This was especially true when the users were looking for some element of text in a large document.  The scanning ability provided to them by IC allowed them to find it much faster.

I feel that systems like this could be tremendously usefull.  I am always behind on my reading partially because I have no time to get any of it done.  The ability to simply scan a document would be tremendously helpful to me.  I feel that the system could become greatly overused however and at times very key elements could be missed causing great losses to some people. 

You have to wonder, if these systems become popular, will they help the users by giving them more free time, or will people, thinking that we can now read more, just throw more documents at us causing us to be back where we started time wise.  Questions like this may make systems like this ineffective in their quest to give the user more free time.  With that being said, however, I would love to see an effective system like this come to frutition.  It seems like I could learn a lot more if I had these capabilities because I would be able to read much more and gain more usefull knowledge.  I would urge the authors to continue their work.