Chris Pollett > Students >
Sandhya
    ( Print View )

    [Bio]

    [Project Blog]

    [CS297 Proposal]

    [Autosuggest-PDF]

    [English dictionary Trie]

    [Google Autosuggest]

    [Autosuggestion in Yioop]

    [Multi-word Autosuggest]

    [CS297 Report-PDF]

    [CS298 Proposal]

    [Autosuggest for foreign languages]

    [CS298 Project Report-PDF]

    [CS298 Presentation-PDF]

    [Project Code-ZIP] 

                          

























Deliverable-2

Objective

To examine and analyse how Google Autosuggest functionality works

Tools used

  • Safari - Activity window
  • Firefox - Webconsole

I coducted series of experiments on Google search page with various kinds of inputs and examined using the Safari Activity window, the actions going in the back end.

The activity window display on opening a google search page looks some thing like below


Activity window
The screen shot below shows various javascript/image and other files loaded while a query is being entered

Activity window
Data loaded after every click can be downloaded using the activity window. The downloaded file has JSON data with the url in hex format and the url contains all the suggested values for the query
Below is the file downloaded for the above search query "Soc"
Downloaded file
The set of urls in the activity window after every letter is entered in "Soc"
https://www.google.com/s?hl=enampgs_nf=1ampcp=1ampgs_id=3ampxhr=tampq=Samppf=pampoutput=searchampsclient=psy-abampoq=ampaq=ampaqi=ampaql=ampgs_l=amppbx=1ampbav=on.2,or.r_gc.r_pw.r_qf.,cf.osbampfp=c536c393c8baad2dampbiw=708ampbih=706amptch=1ampech=1amppsi=8wSXT8b1NMbliALimOndDw.1335297267994.1
https://www.google.com/s?hl=enampgs_nf=1ampcp=2ampgs_id=8ampxhr=tampq=Soamppf=pampoutput=searchampsclient=psy-abampoq=ampaq=ampaqi=ampaql=ampgs_l=amppbx=1ampbav=on.2,or.r_gc.r_pw.r_qf.,cf.osbampfp=c536c393c8baad2dampbiw=708ampbih=706amptch=1ampech=2amppsi=8wSXT8b1NMbliALimOndDw.1335297267994.1
https://www.google.com/s?hl=enampgs_nf=1ampcp=3ampgs_id=campxhr=tampq=Socamppf=pampoutput=searchampsclient=psy-abampoq=ampaq=ampaqi=ampaql=ampgs_l=amppbx=1ampbav=on.2,or.r_gc.r_pw.r_qf.,cf.osbampfp=c536c393c8baad2dampbiw=708ampbih=706amptch=1ampech=3amppsi=8wSXT8b1NMbliALimOndDw.1335297267994.1
Trying to analyze the urls,
nf - is always 1
cp - gives numbering to the urls 1,2,3...
gs_id - Should be Google search Id
tch - is always 1
ech - is also numbering but it will be same if a same query is entered twice duwing one search
The figure below shows the highlighted part where we can see the words that are retrived for the autosuggest list
Downloaded file
As we type the second word in the query, the results are displayed using the top item from the suggestion list. A similar file like above can be downloaded which has the styling and content of the result page

I was not able to reverse engineer the autosuggest process but understood the sequence of events occur during search