How Google Search Engine Works
Download as PPT, PDF5 likes2,649 views
The document provides an in-depth overview of the Google search engine database, detailing metrics used in searches such as URL, size, cached links, and various search term techniques including stemming and the inclusion/exclusion of words. It also describes advanced search operators to enhance query effectiveness, like 'intitle:', 'site:', and filetype restrictions, along with tips for better search results. Additionally, it highlights resources for further information, like Google Guide and Google Librarian Center.
How Google Search Engine Works
- 1. Ark-sol.com Research and Bibliography about Google Search Engine
- 3. Database Google Used •URL, size, date last crawled •Cached link •Pages like this one Approximate # of hits Ads selected by Google based on you search terms Search terms are in bold
- 5. Cached reveals the page as Google found it may differ from the current page Cached exists if a page is full-text indexed About 1 billion pages in Google are not cached Not fully searchable no Cached if a page owner requests not to be cached
- 6. And
- 7. The Fuzzy And only some of the words if a page is “important” words may occur only in link to the page words occur somewhere on the site a page belongs to
- 8. Google stems “when appropriate” Includes plural, singular, past, present tense of words in search Search: school librarian Result: library, librarian, library’s, librarian’s Single word searches aren’t stemmed
- 9. Common or Stop words are ignored No official list from Google Auto-phrasing Searches containing only stop words
- 11. More than 100 factors in the metrics On-the-page metrics Word order matters Word frequency Automatic-phrasing In the title In unique fonts In prominent areas (like lists)
- 12. Off-the-page metrics Words describing the link Links on one site to another are like votes-- PageRank Stuffing the ballot box Reputation of the ‘voting’ page Can’t buy a better PageRank PageRank independent of search terms
- 13. But how do I make my searches better? Ark-sol.com
- 14. + Inclusion operator Force searches on stop words Turns off stemming Use quotation marks for phrases “public librarian” 234,000 .4% of public librarian 58,600,000 Forces searches on stop words Turns off stemming
- 15. Hyphen makes phrases and searches with and without hyphens bite-sized retrieves: bite-sized, bite sized, bitesized Other examples?
- 16. Or Not
- 17. OR search Search for two terms at once - exclusion operator Use with care; Search: twins Minnesota 2,750,000 Eliminate undesired words twins Minnesota –sports 1,300,000
- 18. * full-word wild card, word substitution Ideal for partly remembered quotes Searching for answers to questions Proximity searches ~ synonym operator ~guide searches for: tutorial, manual, help, map, tips
- 19. Intitle: terms are searched for in title only Pages concentrate on term Hybrid cars intitle:mileage Combine with OR intitle:"new urbanism" OR intitle:"sustainable communities” allintitle: Combine with site: allintitle: hybrid cars mileage –site:.com
- 20. Limit to a domain (edu, com, etc) site:edu OR site:gov OR site:lib.co.us Search within a site site:memory.loc.gov “dust bowl” Use Google as a search engine for a site Can ONLY use first part of URL Omit http: & final / inurl:dustbowl searches for term anywhere in URL
- 21. Filetype: Search for a particular type of document tax return filetype:pdf Exclude a filetype -filetype:xls Can use view as HTML Avoid viruses Allows you to read it even if you don’t have the software
- 22. Google Guide http://www.googleguide.com/ Google Librarian Center http://www.google.com/librariancenter/index.html Ark Solutions online marketing http://www.ark-sol.com/online-marketing
Editor's Notes
- #2: REC
- #3: Google doesn’t actually search the web. It searches it’s index of the web… a copy. The doc server assembles the results that the index server produces. This is where Google’s page rank software comes in to determine what order the results are in.
- #4: Stress that Google is searching it’s database of copies of the web, spread out over 500 computers
- #5: Proof that Google is searching a database and not the real web.
- #7: Google’s default, but it’s fuzzy Problems? words can occur anywhere in results pages may have different meanings or contexts some pages may not contain all of your words some may not have any of your words
- #8: And Talk briefly about Boolean searches (how many know what this is)
- #9: Stemming The word is automatically searched as the stem or root with many endings allowed. kite flying retrieves words with kite kites, flying, fly, flyer’s, flyers’, flyers --side note not case sensitive Write in Turn off answers Operator, quotes, single word searches or searches using only ‘stop words’
- #10: Write in Turn off answers Operator Quotes Single word
- #12: Google Metrics Over 100 different factors in each search, algorithm is always changing + spider continually updating database (thus results change) Proprietary software Search words can appear in title of page, link to page, URL of the page & the page itself Pages weight by prominence of words & frequency of words; Searches for all your terms on a page, even better your terms near each other… best of all pages where your search terms appear in the order you typed them. Weights links pointing to the page (popularity contest doesn’t return the most creditable resource) Links from more popular sites are weighted more
- #13: Reputation Some receive high rep by default, gov agencies, well-know or prominent companies, university faculty (smithsonian, nasa, JAMA…) Good rep by association with the above
- #15: Use quotes or inclusion operator to turn off stemming, force search on stop words,
- #16: Always use the hyphen on words that might be hyphenated since it searches both Words are treated as a phrase – simliar to w/ quotes Other examples: asian-american, african-american, mother-in-law, ex-wife, e-mail
- #17: OR Useful when: stemming doesn’t cover the variation your looking for; To cover a common misspelling; For synonyms – parent/guardian; Address apostrophe variations Can also use | instead of OR NOT Not isn’t supported by Google, will use (-) instead
- #19: Wildcard Recently ‘softened’ no need to use more than one asterisk per word -The parachute was invented by * - Vitamin * is good for eyes Ask class for examples ~college ~zoo ~library
- #20: Other uses?
- #21: How would you use this? Google toolbar feature
- #22: How would you use this?