Thursday, January 18, 2007

Module 4 - Tasks (Search Engine Task)

  • Part 1 - Choose your most commonly used internet search engine and do a search for the words: advanced internet users. Record the first hit and number of hits...

Search results for 'advanced internet user'

Task 2 - Using copernic or similar , set it up to search at least three search engines (including one that will search the deep web) and repeat exactly the same search. Record the number of hits ... and compare to your first search. What differences did you notice? Why? Which search, on first glance gave you the most promising results?



CompletePlanet


Search results for 'advanced internet user'

  • Search yielded 301 hits to databases
  • 1st result: All Bidz In - an internet auction site
  • Deepweb search
  • Hits ranked based on Deep Query Manager (DQM) settings.
  • Initial hits demonstrate low relavancy


Ask.com

Search results for 'advanced internet user'

  • Search yielded 24,550,000 hits
  • 1st result: Advanced Users
  • 2nd generation search engine
  • results ranked on number of links from highly rated web pages
  • results clustered by concept/type of site
  • able to retrieve from deep web (i.e. images)
  • initial hits suggest relevance to search


Ixquick



Search results for 'advanced internet user'

  • Search yielded 24,549,846 hits
  • 1st result: bulb link: Internet user education
  • 2nd generation search engine
  • meta (or mega) search engine
  • uses many search engines at the same time, eliminating duplicate files
  • able to access deep web
  • expected to produce most relevant result given meta attributes

I chose the last three engines due to their differences in their retreival and ranking of hits, as illustrated through the bullet points above. These differences are one of several reasons why the results differ so much from the search carried out on Google. Some search engines, particularly metasearches, have a predefined cap on the number of hits taken and the number of hits displayed. This is why Ixquick and Ask.com differ so much from the massive 309 million hits from Google.

Interestingly, despite the differing search methodology utilised with Ixquick and Ask.com both yielded a similar number of hits. Perhaps this suggests there are similar results parameters within their respective programming, despite differing methods of retrieval and ranking. Ixquick could be expected to find the newest material on the Internet, given the time restraints in building ranking through links, such as the case with Google and Ask.com.

It is also noted that a search through the deep web yielded a much smaller number of overall hits. I assume this is due to the fact that each link takes you to a database of information (most of which seem defunct) as opposed to a link to specific 'fixed' knowledge. However, I was still surprised to find only 301 hits. Does CompletePlanet have a results cap or is this all the material it could find within the databases it is able to access?

It may be worth mentioning I initially thought to use Surfwax. Each time I started the search everything crashed and the computer needed a reboot. After three times I wrote the site off as 'never to be seen again!'.

No comments: