Text mining system for web-based business intelligence applied to web site server logs

• US 7,330,850 B1
• Filed: 10/04/2001
• Issued: 02/12/2008
• Est. Priority Date: 10/04/2000
• Status: Expired

• Current Assignee: Inetprofit Com Incorporated, Alto Dynamics, LLC (IP Investments Group LLC)
• Original Assignee: ReachForce, Inc. (American Express Company)
• Inventors: Seibel, John C., Feng, Yu, Foster, Robert L.
• Primary Examiner(s): Fleurantin; Jean Bolte
• Application Number:
  US09/971,334
• Time in Patent Office: 2,322 Days
• Field of Search: 707 1- 10, 707100-1041, 707200-206, 709/216, 709217-219, 709/203, 709/204, 705/10, 705/14, 705/35
• US Class Current: 1/1
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06F 16/958 Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99937 Sorting

Abstract

A text mining system for collecting business intelligence about a client, as well as for identifying prospective customers of the client, for use in a lead generation system accessible by the client via the Internet. The text mining system has various components, including a data acquisition process that extracts textual data from Internet web sites, including their logs, content, processes, and transactions. The system compares log data to content and process data, and relates the results of the comparison to transaction data. This permits the system to provide aggregate cluster data representing statistics useful for customer lead generation.

Claims

1. A text mining system for providing data representing Internet activities of a visitor to a web site of a business enterprise, comprising:

a data acquisition process, operable to;

extract visitor identification data from a server log of the web site, wherein the visitor identification data identifies a visitor to the web site at a known time;

aggregate the visitor identification data with visitor purchase data to provide aggregated visitor data that represents whether a purchase was made from the website by the visitor at or near the known time;

extract text documents from Internet-wide text sources, the Internet-wide text sources selected from the group of;

newsgroups, discussion forums, and mailing lists to provide visitor related documents; and

extract predictive statistics from the aggregated visitor data to provide extracted predictive statistics;

a server, operable to;

receive one or more queries, wherein each query of the one or more queries represents a request for information about the visitor and the visitor related documents; and

provide responses to the one or more queries based on the received one or more queries and the aggregated visitor data, the extracted predictive statistics, and the visitor related documents;

wherein the server is accessible via a web browser over the Internet.

2. The system of claim 1, wherein the text mining server is further operable to generate and store information maps representing the aggregated visitor data and the visitor related documents.

3. A text mining method for providing data representing Internet activities of a visitor to a website of a business enterprise, comprising:

extracting visitor identification data from a server log of the web site, the data identifying a visitor to the website at a known time

aggregating the visitor identification data with visitor purchase data to provide aggregated visitor data that represents whether a purchase was made from the website by the visitor at or near the same time; and

extracting text documents from Internet-wide text sources other than the website, the Internet-wide text sources selected from the group of;

newsgroups, discussion forums, and mailing lists to provide visitor related documents;

extracting predictive statistics based on said extracting visitor identification data, said aggregating, and said extracting the text documents to provide extracted predictive statistics;

receiving one or more queries, wherein each query of the one or more queries represents a request for information about the visitor and the visitor related documents;

generating results based on the one or more queries and the aggregated visitor data, the extracted predictive statistics, and the visitor related documents; and

storing the generated results.

4. The method of claim 3, further comprising:

generating and storing information maps representing the aggregated visitor data and the visitor related documents.

5. A method, comprising:

extracting visitor identification data from a server log of a website of an e-commerce client, wherein the visitor identification data identifies a visitor to the website at a known time;

aggregating the visitor identification data with information related to at least one of web data or processes occurring at or near the known time to generate aggregated visitor data;

determining whether the visitor purchased a product at or near the known time;

storing information regarding the visitor and activity of the visitor based on said determining, operable to be provided to the e-commerce client; and

extracting predictive statistics from the information regarding the visitor and the activity of the visitor to provide extracted predictive statistics, wherein the extracted predictive statistics are operable to be provided to the e-commerce client.

6. The method of claim 5, wherein the predictive statistics comprise one or more of:

profitable aggregation clusters;

least profitable aggregation clusters;

mean aggregation clusters; and

dropped transaction aggregation clusters.

7. The method of claim 5, further comprising:

overlaying the statistics with one or more of transaction, survey, or user-entered demographics and preferences.

8. A method, comprising:

extracting first information comprising visitor identification data from a server log of a website, wherein the visitor identification data identifies a visitor to the website at a known time;

determining second information related to at least one of web data or processes corresponding to the visitor identification data and occurring at or near the known time, wherein the second information corresponds to activity of the visitor;

determining third information regarding whether the visitor purchased a product at or near the known time;

extracting predictive statistics based on the first information, the second information, and the third information to provide extracted predictive statistics;

storing the first information comprising the visitor identification data, the second information comprising the activity of the visitor, the third information regarding visitor purchase in a memory, and the extracted predictive statistics;

wherein the first, second, third information, and extracted predictive statistics are useable to evaluate the website.

9. The method of claim 8, further comprising:

providing the first, second and third information to a requesting entity associated with the website;

the requesting entity adjusting content of the website based on the first, second and third information.

10. A system, comprising:

a server log that stores information regarding visitors to a website;

at least one first server operable to;

extract visitor identification data from the server log of the website, wherein the visitor identification data identifies a visitor to the website at a known time;

determine first information related to at least one of web data or processes corresponding to the visitor identification data and occurring at or near the known time;

determine whether the visitor purchased a product at or near the known time based on the visitor identification data and the first information; and

extract predictive statistics based on information regarding the visitor and activity of the visitor to provide extracted predictive statistics;

wherein the at least one first server comprises a memory operable to store the visitor identification data, the first information, the extracted predictive statistics, and information regarding whether the visitor purchased a product at the known time;

wherein the at least one first server comprises a web server interface accessible by a client web browser to provide statistics regarding visitors to the website.

11. A computer readable memory medium storing program instructions executable by a processor to:

extract first information comprising visitor identification data from a server log of a website, wherein the visitor identification data identifies a visitor to the website at a known time;

determine second information related to at least one of web data or processes corresponding to the visitor identification data and occurring at or near the known time, wherein the second information corresponds to activity of the visitor;

determine third information regarding whether the visitor purchased a product at or near the known time;

store the first information comprising the visitor identification data, the second information comprising the activity of the visitor, and the third information regarding visitor purchase in a memory;

wherein the first, second, and third information are useable to evaluate the website; and

extract predictive statistics from at least one of the first, second, or third information to provide extracted predictive statistics, wherein the extracted predictive statistics are usable to evaluate the website.

12. The method of claim 11, wherein the predictive statistics comprise one or more of:

profitable aggregation clusters;

least profitable aggregation clusters;

mean aggregation clusters;

or

dropped transaction aggregation clusters.

13. The method of claim 11, wherein the program instructions are further executable to:

overlay the statistics with one or more of transaction, survey, or user-entered demographics and preferences.

14. The method of claim 11, wherein the program instructions are further executable to:

extract unstructured text documents from unstructured Internet sources other than the website to provide visitor related documents.

15. The method of claim 11, wherein the unstructured Internet sources comprise one or more of:

newsgroups;

discussion forums; and

mailing lists.