How does a search engine personalise results?

Most popular search engines personalise, whereby past searches influence present and future, results. What are the main ways a search engine personalises results and why do they do this?

Personalisation: How and why?

Google decided to personalise search as a trail on 29th March 2004. So personalisation is not a new feature.

From an information retrieval viewpoint personalisation improves retrieval effectiveness by adjusting search results based on a searcher’s interests. These interests can be observed by mining short-term or long-term behaviours.

Short term
Search engine mines current search session.
Long term
Searches from many previous sessions are mined.

Personalisation ultimately occurs to everyone but it can be reduced by making smart use of technologies, for example, incognito mode or using a cookie removing facility.

VuGraph personalisation.

VuGraph personalisation.

Browsing & search history

Browsing history, search history and user’s explicitly declaring their interests to a search engine all help build interest profiles. Some studies do not take a broad overview of the various technologies. This means that smaller, more obscure, technologies are simply ignored, such as, cookies, for instance. It is possible for cookies to store keywords which can help search engines to personalise results.

Read more about other personalisation technologies.

Posted by

References

  1. Hines, M. (2004) Google takes searching personally. [Online] [Accessed on 13th August 2013]
  2. Shen, X. Tan, B. Zhai, C. (2005) Context-sensitive information retrieval using implicit feedback. SIGIR. pp. 43 — 50
  3. Shen, X. Tan, B. Zhai, C. (2006) Mining long-term search history to improve search accuracy. KDD. pp. 718 — 723
  4. Sontag, D. Collins-Thompson, K. Bennett, P.N. White, R.W. Dumais, S.T. and von Billerback, B. (2012) Probabilistic models for personalizing web search. WSDM. pp. 433 — 442
  5. White, R.W. Bennett, P.N. and Dumais, S.T. (2010) Predicting short-term interests using activity-based search context. CIKM. pp. 1009 — 1018
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s