What makes Google so special?. Review on Google

What makes Google so special?

MS restricts me to write a review less than 8000 characters and does not permit me to write two reviews on same topic. Hence I will only mention the benefits and recommend members to read Asyis article on Google.

Some benefits are:

1 BILLION PLUS URLs : Most search engines cover 10 million, 50 million, 200 million but Google is probably only one that can boast of index comprising of more than 1 billion URLs.

SPEED : There are amazing timings recorded. The bar on top shows results like 1-10 of 589 in 0.04 secs. How? Wow!

THE RESULTS ON TOP YIELD THE CORRECT INFORMATION : As if speed wasn’t enough to awestruck me, this feature is like a real “Google” – maybe Kumble knows about it? But its true! Unlike most other search engines, Google only produces results that match all the search terms. This may be either in the link or some text on the Web page. Hence the user does not land in problems mentioned above. But there’s more to it. Besides displaying only the results that have all the search terms, Google also analyzes the proximity of the search terms within the page and displays those results first where the terms are closest to each other and in the order specified. Even the position of search terms is treated with respect.

i.e for above example, the first few result will likely be

Green House Effect: Global warming…

Green House Effect blamed for the irregular weather patterns….

This in technical term is called “relevance”. How do they achieve this in 0.04 sec? Start wondering now….

CACHE ALLOWED: As Google crawls the web, it saves the snapshot of the web page, which is available through a link “cached” below each search result. The advantage here is that if the original link is not available due to server problems, the web page may be accessed through this snapshot. Of course this page may not be recent or updated, but if it is not a news item (time-sensitive) you are searching, then the purpose is solved.

NO DEAD LINKS: Google crawls the web quite often and updates its database frequently thus eliminating dead links.

LOOK BEFORE YOU LEAP : Some web pages may have a summary which most often is not changed though the web page contents are changed. So instead of showing such summaries, Google shows an excerpt of the text that matches the query with search terms highlighted in bold. This little concept is quite thoughtful and helps in choosing whether this particular link should be clicked or that one seems more appropriate.

All the benefits are understood but then how does Google achieve it?

Earlier versions of search engines needed you to advertise or register yourself with them. Others (even now) search for keywords, which are present in “meta tags” usually in header information in html page. The sorting method is referred as “collection frequency weighting”. The method returns thousands of results, giving more weight to pages in which the keyword appears in the title, at the top of page or in meta tag. There are other disadvantages besides the irrelevant results such as “spamming” i.e. any web author may just enter all important and hot keywords in meta tag irrespective of whether the page has that content or no.

Google is the first one to use a different approach. It weighs a page based on the number of links that point to that page. Say for eg. All the members on this site have their own web page and each of them have a link to my web page then my web page will be ranked higher. The rational behind this is that the author of web page has nothing to do with its own ranking and authors of different web page must have given some thought before putting up the link to a particular page.

The second aspect is that if the inbound link is coming from a web page that is itself highly rated then the importance of the page will be higher as compared to some page that may have several inbound links from not so valued pages. For eg: If Yahoo homepage or Amazon homepage that are highly rated have links to Faisal’s web page, he will have higher ranking as compared to my web page where several members have links to.

This is a celebrated PageRankTM technology that gave Google the accolades and the popularity. Google has several pcs connected together that crawl the web reading 50 million pages per day and indexing them and extracting the URLs. If URL is new then it is weighted as mentioned above and given some rank before being stored in database. Also web pages are stored in a database as snapshots and made available through “cache”. If it is found to already exist in database then the cache is updated.

Any flipside?

FLAW IN PAGERANK:

Although the above-mentioned technology is a significant improvement over traditional ranking systems, there are many instances where it does not necessarily address the question, “Is this page really delivering what you’re looking for?” Complex?

Let me explain. PageRank technology looks only at the connectivity graph of the link structures, completely ignoring the context of those links. It only tells that this page is important but it may not be relevant. Eg. If you are looking for nuclear physics and I have a paragraph on “nuclear physics” and Yahoo! has a link to my page, it is likely to be listed first in search results because it is important but is Yahoo! the ultimate authority in nuclear physics or would you rather trust BARC or TIFR or ISRO for that matter.

UNFRIENDLY FOR WEB DESIGNERS

Imagine you as a dynamic web designer making frequent changes to your web page to support the latest news or technology and I searching for some information and choosing the “cache” link that takes me to your page as it was 5 weeks before. Is there any point in you updating your page? Besides what about copyright issues when 100% of your content is cached by Google servers?

Secondly web crawlers are supposed to abide by rules set in a file “\robots.txt” which has rights to certain pages or denial of permission to index certain pages. However, Google occasionally indexes web pages even if the robots.txt prohibits the listing of pages. Worse is that if you block their robots through this file, there is grave danger that your page will never be crawled again, even if you re-submit the page information.

NO ADVERTISEMENTS? THINK AGAIN!

Many claim that the site is devoid of graphics. Good – agreeable. There are no advertisements. What??? Come again. Look more carefully! How do you explain that text box on you search result page that is displayed first? OK a little text box doesn’t hurt. But lets look at it from another angle – what if it is my own text box being displayed. A natural beaming face is expected from me, wouldn’t you agree? But how appealing are text ads? What are chances that people desperately seeking information on Google will click my text ad?

But all said and done, as of today Google is the best search engine, but ….. for how long? Is there any successor? Wait till my next review….