The future of search engines

The technology adopted by current search engines is still considered as a new technology. Was introduced in 1998 as an advancement towards the old search methods, the web search today is still a very young technology with a great opportunity for growth. This essay will cover some indications for future growth of the current search engines.

Clustered search results
Most search engines available on internet nowadays are still not able to overcome the problems of keyword synonymy and polysemy [1]. For example, imagine that there is a user who wishes to find information about Cocomo beach located in Fiji island and enters the keyword ‘cocomo’ into a search engine. This unfortunate user will certainly be jolted with plenty of less relevant web-pages such as pages about Cost Constructive Model which is a model used for developing software in IT. 

One possible way to deal with this problem is by attaching the search engines with a clustering facility for rendering search results. In other words, the results displayed by these search engines will be grouped based on their similarity. With this approach, search engine users will be allowed to ‘drill down’ the result list and find the category that suits them the most [2]. Therefore, in the cocomo example mentioned above, this user will be able to focus his research only on the pages displayed under ‘cocomo beach’ category.

  
Document visualization

One major challenge faced by current search engines is to overcome the ‘information overload’ problem in rendering their search results [3]. One alternative way proposed for solving this problem is by summarizing this large amount of information into two-dimensional display format such as maps and icons. The main objective of this approach is to provide internet users with advanced user-friendly search facilities [2].  There is one search engine hosted in Switzerland that already adopted this type of technology and won the Best of the Swiss Web Prize [2]. This search engine, search.ch, displays icons of restaurants, movie theatres, trams, etc acts like a complete map of Switzerland [2]. By using icons such as icons of restaurants or movie theatres for determining the pages displayed in search results, users are expected to have better experience while searching on internet.

Result of a research conducted in 2005 showed that respondents who experienced document visualization in web searching indicated the positive opinions or ‘subjective ratings’ [4].  The research used three different shapes of icons [4] to show the relevancies of pages displayed in search results. These icons appeared at the left hand side of search result lists to indicate number of keywords occurrences in a page [4]. So, as a future development plan, other popular search engines may consider to adopt this technique or bring this technology into more advanced level, such as to cover a wider search area instead of focusing only on the geographical map of a country.

Personalization features
Some search engines nowadays are providing personalization facilities to internet users. Search engines such as A9 (www.a9.com) and Google Web History display and rank their search results tailored to user’s taste and preferences. By using the information such as what pages that have ever been visited by a user, along with the frequency and the attempted queries in the past, these search engines are capable for making recommendations based on user preferences [2]. 

There is one method suggested by Srour et. al aiming for enhancing the performance of search engine personalization. This method measures the level of trust and similarity between users and then generates a personalized search ranking based on this calculation [5]. So, in addition to analyzing the user’s search history, the search engine will require a user to submit his profile containing details on his user name, password, occupation and preferences [5]. This user is then required to recommend a person that he can trust based on the assumption that this person will have the similar characteristics and preferences with the user himself [5]. Therefore, by using the acquired information, a search engine can now personalize its search results tailored to user preferences.

Time-sensitive searches
Some information searched by internet user is time-sensitive so that pages published at different time will bring different value. Information such as news articles will probably not be valued by it readers once the articles have become obsolete. For example, an article that talks about the 9/11 tragedy written in 2001 will certainly has a different content from the ones that were just recently posted on internet [2].
Search engines such as Google Scholar enable their users to range the publishing years of the pages they want to retrieve. Other search engines that specialized in retrieving news articles such as Allinonenews.com even provide information about the publishing date of each article that they found [6]. So, in case that other popular search engines start to provide these abilities in the future, users on internet will probably not face much difficulty in conducting time-sensitive searches again in the future.

 

 

[1] Shepherd, S.J. (2007). Concepts and Architectures for Next-Generation Information Search Engines. International Journal of Information Management, 27:1, p. 3-8. http://www.sciencedirect.com/science?ob=ArticleURL&_udi=B6VB4-4MW90DT-1&_user=554534&_coverDate=02%2F28%2F2007&_alid=737300299&_rdoc=4&_fmt=high&_orig=search&_cdi=5916&_sort=d&_st=4&_docanchor=&_ct=4&_acct=C000028338&_version=1&_urlVersion=0&_userid=554534&md5=ec334c8b7f3202bcf89e1ade9e0f62f0

[2] Langville, A.N. and Carl D. Meyer (2006). ,Google’s Page Rank and Beyond: The Science of Search Engine Rankings, Princeton University Press, New Jersey

[3] Wang, Y., Dai, W. and YuFei Yuan. (2007). Website browsing aid: A navigation graph-based recommendation system, Decision Support Systems, http://www.sciencedirect.com/science/article/B6V8S-4NS2G9M-1/1/938603da6c1951b1a1e09ac133d5fdbc

[4] Heimonen, T. and Natalie Jhaveri (2005), Visualizing query occurrence in search result lists, Information Visualisation, pp. 877-882, http://ieeexplore.ieee.org/iel5/10086/32319/01509175.pdf?isnumber=32319&prod=STD&arnumber=1509175&arnumber=1509175&arSt=+877&ared=+882&arAuthor=Heimonen%2C+T.%3B+Jhaveri%2C+N.

[5] Srour, L., Kayssi, A. and Ali Chehab. (2007), Personalized Web Page Ranking Using Trust and Similarity. Computer Systems and Applications, p. 454-457, http://ieeexplore.ieee.org/iel5/4230920/4230921/04230996.pdf?tp=&isnumber=4230921&arnumber=4230996&punumber=4230920

[6] Lu, Y., Meng,W., Zhang,W., Liu, K. and Clement Yu (2006), Automatic Extraction of Publication Time from News Search Results, Data Engineering Workshops. p.50-50, http://ieeexplore.ieee.org/iel5/10810/34089/01623845.pdf?isnumber=34089&prod=STD&arnumber=1623845&arnumber=1623845&arSt=50&ared=50&arAuthor=Yiyao+Lu%3B+Weiyi+Meng%3B+Wanjing+Zhang%3B+King-Lup+Liu%3B+Clement+Yu

 

Other related topics:

What is E-Business?
How does e-business earn money online?
What makes a good website?
What is CRM? How can it affect on e-business?
Linkspam: What is that and how to deal with it?
Optimizing your e-business with SEO
Search engine: how does it work?