Apparatus and method for automatic assignment of industry classification codes

• US 8,965,877 B2
• Filed: 03/14/2013
• Issued: 02/24/2015
• Est. Priority Date: 03/14/2013
• Status: Active Grant

• Current Assignee: Networks Glenbrook
• Original Assignee: Networks Glenbrook
• Inventors: Komissarchik, Julia, Komissarchik, Edward
• Primary Examiner(s): Truong, Dennis
• Application Number:
  US13/828,587
• Publication Number:
  US 20140280014A1
• Time in Patent Office: 712 Days
• Field of Search: 707/602, 707/710, 707/736, 707/737, 707/740, 706/12
• US Class Current: 707/710
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling

Abstract

A system and methods for automatically assigning of classification codes to a business based on information about the business collected from the Internet are provided in which data extracted from trawling the Internet is compared to a node structure based on a taxonomy of a selected business classification code system.

Claims

1. A system for automatically assigning a business classification code to a company, comprising:

one or more processing units configured to;

trawl the Internet, locate and extract web data relevant to the company;

generate a business classification analysis with a node structure corresponding to a selected business classification code system and to compute a taxonomy word histogram based on the node structure;

generate an extracted word histogram corresponding to the presence of selected web data elements within the extracted web data relevant to the company; and

determine a business classification code assignment with a first list of matches for the business classification code for the company by comparing a normalized scalar product of the taxonomy word histogram and the extracted word histogram to a predetermined threshold.

2. The system of claim 1, wherein the one or more processing units are further configured to compute a TF-IDF distance for each taxonomy node and entity description to determine a second list of matches for the business classification code for the company.

3. The system of claim 2, wherein the one or more processing units are further configured to combine the first list of matches and the second list of matches to determine and output a list of N-best matches for the business classification code for the company.

4. The system of claim 1, wherein the selected business classification code system SIC or NAICS.

5. The system of claim 1, wherein the node structure comprises n-grams of descriptions corresponding to descriptions employed within the selected business classification code taxonomy.

6. The system of claim 1, wherein the system comprises a web trawling component, a web page analysis component and a fact extraction component.

7. The system of claim 1, wherein the one or more processors are configured to locate and extract web data relevant to the company including the company name, company history, product name and product description.

8. The system of claim 1, wherein the one or more processing units are configured to trawl websites including at least one of a company website, a governmental body website, an industry analyst website, a social networking website and a review website.

9. A non-transitory computer-readable storage medium having instructions stored thereon, which, when executed by one or more processors of one or more computing devices, cause the one or more processors to perform operations comprising:

trawling the Internet to locate and extract data relevant to a company;

generating a node structure corresponding to a selected business classification code system and computing a taxonomy word histogram based on the node structure;

generating an extracted word histogram corresponding to the presence of selected data elements within the extracted data relevant to the company; and

determining a first list of matches for the business classification code for the company by comparing a normalized scalar product of the taxonomy word histogram and the extracted word histogram to a predetermined threshold.

10. The non-transitory computer-readable storage medium of claim 9, further comprising computing a TF-IDF distance for each taxonomy node and entity description to determine a second list of matches for the business classification code for the company.

11. The non-transitory computer-readable storage medium of claim 10, further comprising combining the first list of matches and the second list of matches to determine and output a list of N-best matches for the business classification code for the company.

12. The non-transitory computer-readable storage medium of claim 9, wherein the selected business classification code system SIC or NAICS.

13. The non-transitory computer-readable storage medium of claim 9, wherein the node structure is generated by concatenating successive subclasses of descriptions employed within the selected business classification code taxonomy.

14. The non-transitory computer-readable storage medium of claim 9, wherein data extracted relevant to the company includes the company name, company history, product name and product description.

15. The non-transitory computer-readable storage medium of claim 9, wherein trawling the Internet to locate and extract data relevant to the company comprises trawling at least one of a company website, a governmental body website, an industry analyst website, a social networking website and a review website.