Format

GCA takes a request formatted in JSON. The content of the request depends on the corpus queried. You can use the examples provided in the folder RequestSamples.

Here is what a request for Google Books looks like (this is the request used for the wiki Home example on the virtual internet peak ):

{    
	"Request": {    
   	 	"Corpus": "books",    
    	"YearDebut": "1990",    
    	"YearEnd": "2011",    
    	"TimeInterval": "1",    
    	"Language": "en",    
    	"NullThreshold": "100",    
    	"Expressions": [    
        	{    
            	"Expression": "internet"     
        	},    
        	{    
            	"Expression": "internet virtual"    
        	}    
    	]    
	}    
}    

Parameters vary from corpus to corpus (books, patents, scholar...), but a set of common parameters is time-series

Time-series

Queries are only serialized by years for now (more granularity will be possible in the future). Time-series take 3 parameters, a debut, and end, and an interval:

• YearDebut
• YearEnd
• Interval (in years)

For instance:

"YearDebut": "1900",    
"YearEnd": "2000",    
"TimeInterval": "10",

In this case GCA will make 10 queries, one per decade (every 10 years), from 1900 to 2000.

Parameters and examples for specific corpora

• Books
• Patents