The idea of personalization isn’t new; it’s a concept that’s been around for years and has been attracting much attention recently. Personalized search is a promising way to improve the accuracy of web search; however, it often raises serious user concerns.
One of the major deficiencies is that the search is temporal. As such, critics point out that, just because people search for something doesn’t mean that they are genuinely interested in it. That said, the question is: How long will it take before the search engine stops displaying results based on something we user’s were only temporarily interested in. With no clear answer, this is one of the first obstacles that must be overcome for this search to work.
Ambiguous terms pose another obstacle. Words commonly have more then one meaning, thus a single query may mean different things for different users. For example when you search for “apple”, do you mean the fruit or the computer? When you search for “Saturn”, do you want to find the information about the planet or the car? To be successful, search engines will need to learn how to differentiate and determine which terms the user means.
Privacy concerns also pose an obstacle for the personalized search. Because effective personalized search requires collecting and aggregating user information, it raises serious privacy concerns for many users. People don’t want to reveal their personal information and they don’t trust the engines or what these engines may do with it. Thus there appears to be a dilemma between providing high-accuracy web search and privacy preservation. Indeed, the privacy concern is one of the major barriers in deploying serious personalized search applications. As a solution, the search engines should make the value of personalization explicit to the searcher. If the searchers believe that they are getting real value for giving up their privacy, they may do so willingly.
The next major concern deals with identification of individual users. How will the search engine know who is who? This is essential when multiple users share a computer or when users switch locations. Without the ability to identify and authenticate users automatically and adapting to their needs, personalized search will not be adding value, but on the contrary, disrupt the user’s search. Moreover, people often search for others. Personalization would not only be useless in this case, it will be detrimental.
Personalized search is still in its early stages, but it is a great tool of the future. There are a lot of possibilities and potential to be found in personalization, and there are areas which still need work before the search is really user friendly, provides better results and a more personal, effective search experience.