0

I am testing AlchemyAPI for extracting the name of company from a given website. I am able to get the correct information in most of the cases. I also need to categorize the result into a business. What I get from AlchemyAPI is the most general classification but I need a more specific answer. For example, in their demo link, if I input www.ford.com it returns the category as recreation. DMOZ also classifies ford as recreation at the top most level. I need to classify Ford as Business-vehicles. Is there any way I can do this? Does AlchemyAPI use DMOZ to classify and if so, how to achieve a more specific classification?

Another example is Pepsi. Pepsico is classified as business. I need to classify it as Business --> Food and Beverages.

mpvyas
  • 1
  • 1

2 Answers2

0

It sounds like you are using AlchemyAPI's text categorization endpoint. That call will attempt to categorize the entire document into 1 of 12 categories ("business" and "recreation" being two of them).

The text categorization call is intended to say "this document is generally about X". If you want to extract specific company names from a document and figure out the category the company is in, try using the entity extraction call (More info at: http://www.alchemyapi.com/api/entity-extraction/). That call will find all of the companies in the document, and it will return a "type" for each company. Instead of the 12 categories in the other call, the entity extraction call supports hundreds of types. That should get you more precision for classification.

Steve
  • 141
  • 1
  • 3
0

AlchemyAPI's ranked named entity extraction will categorize a named entity and if you enable disambiguation, it will give you a more specific categorization (i.e. Ford = Company -> AutomobileCompany).

mageeky90
  • 139
  • 1
  • 4