Some embodiments provide a method for identifying business lines of a company. The method classifies several documents as relevant to several different business lines and several different companies. For a particular company and particular business line, the method identifies a number of documents classified as relevant to both the particular company and the particular business line. When the identified number of documents exceeds a particular threshold, the method associates the particular business line as a business line of the particular company. In some embodiments, the method calculates a score for each business line in the set. The score for a particular business line represents the importance of the particular business line to the particular company. The method sorts the business lines in the set based on the calculated scores.
展开▼
