Not hidden but less well known facts on Google is on its power of knowledge on Internet and user activities. This project will investigate the use of Google specifically on the fact not very open to the public. I will try to come up with simple test applications to demonstrate ideas.

Screen shots

Background

The current Google interface is exactly for text information retrieval. A user puts some keywords composed in Boolean algebra and got the list of found results. Google puts some ads to make a fortune. Hence in the view of Google customers, who actually pay money to Google, keywords are the most and all the most important things to ask Google to link their company homepage. So there is a concept of auction for customers to buy popular keywords. To help(?) customers, Google provides the tool to review the current popular statistics at Google AdWords). Yes, this is the rule of game in Web advertisement. At the time when I am writing this article surely Olympic is the most popular one. Imagine 1.5 billion Chinese people plus other guys in the world including myself connected to the web are typing the word Olympic at the Google.

Away from the business point, Google displays the simple statistics at the search result page which is displayed at the top-right corner of the web page under the Google banner in a form like for an instance Results 1 - 100 of about 605,000. So 605,000 means the search keyword approximately exist at 605,000 web pages. However, if you press the next button, the actual page number will never go that much since Google will drop tons from them and only shows the most relevant pages (this is what they say but I do not know the way to filter such non-relevant pages). Anyway, this statistics is sometimes useful in several aspects:

  1. We can compare between keywords which one is most popular, i.e. we may use this number as a unit of popularity.
  2. At least, for internationals including me who use English as their second language, we can get some ideas from this popularity on what we are writing is awkward or not. Well yes, this is why I am developing this application when I am frustrated on the quality of my writing that passed the MS-Word English grammar test though.

Implementation

Google provides an Ajax interface that returns the JSON coded results for Flash and other Non-JavaScript Environments [1]. In JSON format, result.cursor.estimatedResultCount is what we are interested int. The query address for Ajax is slightly different based on the type of search which Google calls Searcher as shown below.

For instance, if we want to get the estimated result count of a keyword Bush, the Google Ajax address is:

http://ajax.googleapis.com/ajax/services/search/web?v=1.0&rsz=small&q=Bush

Try the above link and search estimatedResultCount. This will give you a clear idea for implementation. I will skip all the rest on implementing GUI for this interface since they are shown well in java scripts that you can identify at the demo page [2].

Conclusion

Though I developed this just for fun. Some people like Google customers or researchers in natural language processing may have some interests for eGoogle extension. Please feel free to contact me.

I have several other ideas to add more features to eGoogle. Let me keep them for future updates.

Demo page

Try it! eGoogle

Feedback

Want to leave a comment or have any questions? <comments />