Recipe creation using text analytics

Abstract

Embodiments presented herein provide techniques for generating a recipe using text analytics methods. A server application receives a search query for a specified dish (e.g., from a client computer). The application retrieves recipes relevant to the specified dish (e.g., by crawling websites on the Internet) and extracts text from each recipe. The application performs a variety of text analytics methods on the text to identify different ingredients and preparation methods in creating the dish. Additional analytics are performed to loosely predict a recipe that conforms to the search query as well as dietary restrictions of a user.

What is claimed is: 
     
         1 . A computer-implemented method, comprising:
 receiving, via a processor, a recipe request, wherein the recipe request specifies a dish and a set of criteria;   retrieving a plurality of recipe texts corresponding to the specified dish;   parsing each of the recipe texts to identify one or more ingredients and one or more methods used in preparing the specified dish and to associate metadata to one or more of the identified ingredients and methods;   generating, from the associated metadata of the identified ingredients and methods, a statistical model for preparing the specified dish; and   generating a combination of the identified ingredients and methods for preparing the specified dish based on the specified set of criteria and the statistical model.   
     
     
         2 . The method of  claim 1 , wherein the plurality of recipe texts are retrieved from online recipe websites and local databases. 
     
     
         3 . The method of  claim 1 , wherein the set of criteria includes nutritional information, preferred ingredients, and ingredients to exclude from the generated combination. 
     
     
         4 . The method of  claim 3 , wherein the set of criteria further includes medical information and dietary restrictions associated with a user. 
     
     
         5 . The method of  claim 1 , further comprising:
 returning at least one of the identified ingredients and methods to a client computer, wherein the client computer sent the recipe request;   receiving a selection one or more of the at least one of the identified ingredients and methods; and   generating a combination of the identified ingredients and methods based on the selection.   
     
     
         6 . The method of  claim 1 , wherein the metadata associated with the identified ingredients includes a type, nutritional information, and whether the ingredient is common to the specified recipe, and wherein the metadata associated with the identified methods includes cooking times and preparation temperatures. 
     
     
         7 . The method of  claim 1 , wherein the ingredients and methods are identified based on a frequency of appearances in the parsed recipe texts.