SES NY | Introduction to Information Retrieval on the Web

By Lauren Kade | Mar 22, 2011
More Articles by Lauren

After a delicious networking lunch, I’m now attending “Introduction to Information Retrieval on the Web” with Mike Grehan, Chair of the SES Advisory Board.  This sounds like a fascinating session about the ins and outs of how search engine crawlers gather and index information.

Information Retrieval vs. Data Retrieval

  • Data retrieval provides a solution to a user of a database solution.
  • An information retrieval system must interpret the document contents and rank them according to relevance.

How Search Engines Work

  • Crawlers follow links and collect text.
  • They only have a very small influence over page ranking.
  • It’s possible to achieve a #1 ranking in Google without having your site crawled.

What Crawlers are Looking For in Your Page:

  • keywords in title
  • keywords in headline
  • keywords in body copy
  • keywords in alt tags
  • text links and/or link to site map

Today, search engines are more interested in what other people think of your website (i.e. how many links point to your site).  But it’s not only about quantity- it’s about quality.

Link analysis can uncover “cyber communities,” or pages that are in a specific narrow field of interest.  Links from within your own community are high quality links, and can help you expand your link building.  If you can find the “pope,” or the biggest influencer in your cyber community, and can get a link from their site to yours, that’s the best quality link!

Co-citation: if page C links to both A and B, yet A and B are not linked to each other, if enough pages link to both A and B, search engines will learn that A and B are related.

Changes in End User Behavior

With the latest algorithm updates, images and video are showing higher in the listings.  These media types are eye-catching, and receive more clicks.  People are also interacting almost constantly today, 24-7; this is called connected marketing.  Decision makers no longer act independently of each other but are all the more connected to other consumers.

Understanding User Intent

User intent on search engines can be either:

  • Informational (i.e., Users are researching information).
  • Navigational (i.e., Users know where they want to go and are navigating towards it).
  • Transactional (i.e., The user wants to do something, like shopping.  They want a return which satisfies their need).

Query Chains

Google follows a series of user queries to learn what he actually wanted.  For example, if a user searches for “special edition” first, then realizes the results are not what he wanted, he will research using a different query.  At the end of his multiple tries, the user searches using the query “limited edition” which has the results he was looking for.  The next time someone else searches “special edition,” Google will give them some “limited edition” results, after learning what people who search for “special edition” want.

The Future of Search

  • More user generated content
  • Richer search experience
  • Demand for many types of information
  • Expanded feedback loops

Share this article

Facebook Icon Twitter Icon LinkedIn Icon Google Plus Icon Pinterest Icon

Subscribe today!

Join over 4,000 marketers who receive actionable digital marketing insights.


Blog Search