System and method for adaptive text recommendation

• US 8,645,389 B2
• Filed: 12/03/2004
• Issued: 02/04/2014
• Est. Priority Date: 11/27/2000
• Status: Expired due to Fees

• Current Assignee: Quest Software, Inc.
• Original Assignee: SonicWALL, Inc. (SonicWall Holdings Ltd.)
• Inventors: Oliver, Jonathan James, Buntine, Wray Lindsay, Roumeliotis, George
• Primary Examiner(s): Ahn, Sangwoo
• Application Number:
  US11/003,920
• Publication Number:
  US 20090089272A1
• Time in Patent Office: 3,350 Days
• Field of Search: None
• US Class Current: 707/748
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: H04L 51/18 Commands or executable codes
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

Network system provides a real-time adaptive recommendation set of documents with a high statistical measure of relevancy to the requestor device. The recommendation set is optimized based on analyzing the text of documents of the interest set, categorizing these documents into clusters, extracting keywords representing the themes or concepts of documents in the clusters, and filtering a population of eligible documents accessible to the system utilizing site and or Internet-wide search engines. The system is either automatically or manually invoked and it develops and presents the recommendation set in real-time; for example, upon logging onto a web site or as the client views additional documents or pages of a website. The recommendation set may be presented as a greeting, notification, alert, HTML fragment, fax, voicemail, or automatic classification or routing of customer e-mail, personal e-mail, job postings, and offers for sale or exchange.

Claims

1. A method for adaptive information recommendation, the method comprising:

storing user-specific information in memory, the user-specific information concerning user activity with a plurality of documents; and

executing instructions stored in memory, wherein execution of the instructions by a processor;

clusters an interest set of documents associated with the user activity into one or more clusters, wherein clustering an interest set of documents comprises;

assembling the interest set of documents,

pre-processing words of the interest set of documents, and

grouping documents from the interest set of documents into the clusters utilizing a clustering algorithm that maximizes a cluster score of the clusters, wherein the cluster score is an average similarity score between the documents in the cluster,

identifies a keyword for a cluster of the one or more clusters, the keyword identified based on natural language input by a user and representing the theme of the documents in the cluster,

identifies a set of eligible documents within the cluster of the one or more clusters, each identified document containing either the keyword or the natural language input by the user representing the theme of the documents,

filters the set of eligible documents in the cluster to meet an application criterion, the application criterion based on the user-specific information stored in memory and a user-defined limit on document age, wherein filtering documents does not require user interaction, and

adaptively constructs a recommended set of documents for the cluster from the filtered set of eligible documents based on relevance to the keyword or the natural language input wherein constructing the recommended set of document includes;

calculating a relevance score of each document in the filtered set of eligible documents, wherein the relevance score is based on a number of times the keyword or the natural language input by the user representing the theme appears in each document in the filtered set of eligible documents,

selecting documents of the filtered set of eligible documents with high relevance scores, and

applying a selection criterion measuring popularity of the document in the filtered set of eligible documents.

2. The method of claim 1, wherein the interest set of documents comprises a document previously accessed by the user.

3. The method of claim 1, wherein identifying a keyword for the one or more clusters includes calculating a plurality of keyword scores corresponding to the one or more clusters and selecting the keyword that maximizes the keyword score of the cluster, wherein calculating a keyword score is based on a frequency of a keyword in the interest set and a frequency of the keyword in the cluster.

4. The method of claim 1, wherein the interest set of documents is an interest set of offer descriptions, the keyword represents the theme of the offer description, the eligible set of documents is an eligible set of offer descriptions, and the recommended set of documents is a recommended set of offer descriptions.

5. A non-transitory computer-readable storage medium having embodied thereon a program, the program being executable by a processor to perform a method for adaptive information recommendation, the method comprising:

storing user-specific information in memory, the user-specific information concerning user activity with a plurality of documents;

clustering an interest set of documents associated with the user activity into one or more clusters, wherein clustering an interest set of documents comprises;

assembling the interest set of documents,

pre-processing words of the interest set of documents, and

grouping documents from the interest set of documents into the clusters utilizing a clustering algorithm that maximizes a cluster score of the clusters, wherein the cluster score is an average similarity score between the documents in the cluster;

identifying a keyword for a cluster of the one or more clusters, the keyword identified based on natural language input by the user and representing the theme of the documents in the cluster;

identifying a set of eligible documents within the cluster of the one or more clusters, each identified document containing either the keyword or the natural language input by the user representing the theme of the documents;

filtering the set of eligible documents in the cluster to meet an application criterion, the application criterion based on the user-specific information stored in memory and a user-defined limit on document age, wherein filtering documents does not require user interaction; and

adaptively constructing a recommended set of documents for the cluster from the filtered set of eligible documents based on relevance to the keyword or the natural language input, wherein constructing the recommended set of document includes;

calculating a relevance score of each document in the filtered set of eligible documents, wherein the relevance score is based on a number of times the keyword or the natural language input by the user representing the theme appears in each document in the filtered set of eligible documents,

selecting documents of the filtered set of eligible documents with high relevance scores, and

applying a selection criterion measuring popularity of the document in the filtered set of eligible documents.

6. The non-transitory computer-readable storage medium of claim 5, wherein the interest set of documents comprises a document previously accessed by the user.

7. The non-transitory computer-readable storage medium of claim 5, wherein identifying a keyword for the one or more clusters includes calculating a plurality of keyword scores corresponding to the one or more clusters and selecting the keyword that maximizes the keyword score of the cluster, wherein calculating a keyword score is based on a frequency of a keyword in the interest set and a frequency of the keyword in the cluster.

8. The non-transitory computer-readable storage medium of claim 5, wherein the interest set of documents is an interest set of offer descriptions, the keyword represents the theme of the offer description, the eligible set of documents is an eligible set of offer descriptions, and the recommended set of documents is a recommended set of offer descriptions.

9. An apparatus for providing adaptive information recommendations, the apparatus comprising:

a memory configured to store user-specific information concerning user activity with a plurality of documents; and

a processor configured to execute instructions stored in memory, wherein execution of the instructions by the processor;

clusters an interest set of documents associated with the user activity into the cluster, wherein clustering an interest set of documents comprises;

assembling the interest set of documents,

pre-processing words of the interest set of documents, and

grouping documents from the interest set of documents into the clusters utilizing a clustering algorithm that maximizes a cluster score of the clusters, wherein the cluster score is an average similarity score between the documents in the cluster,

identifies a keyword for a cluster of the one or more clusters, the keyword identified based on natural language input by the user and representing the theme of the documents in the cluster,

identifies a set of eligible documents within the cluster of the one or more clusters, each identified document containing either the keyword or the natural language input by the user representing the theme of the documents,

filters the set of eligible documents in the cluster to meet an application criterion, the application criterion based on the user-specific information stored in memory and a user-defined limit on document age, wherein filtering documents does not require user interaction, and

adaptively constructs a recommended set of documents for the cluster from the filtered set of eligible documents based on relevance to the extracted keyword or the natural language input, wherein constructing the recommended set of document includes;

calculating a relevance score of each document in the filtered set of eligible documents, wherein the relevance score is based on a number of times the keyword or the natural language input by the user representing the theme appears in each document in the filtered set of eligible documents,

selecting documents of the filtered set of eligible documents with high relevance scores, and

applying a selection criterion measuring popularity of the document in the filtered set of eligible documents.