Method and system for probabilistically quantifying and visualizing relevance between two or more citationally or contextually related data objects

• US 20070073748A1
• Filed: 09/27/2005
• Published: 03/29/2007
• Est. Priority Date: 09/27/2005
• Status: Active Grant

• Current Assignee: PatentRatings, LLC
• Original Assignee: PatentRatings, LLC
• Inventors: Barney, Jonathan
• Granted Patent: US 7,716,226 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/14 Details of searching files based on file metadata
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/2465 Query processing support for facilitating data mining operations in structured databases
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/26 Visual data mining; Browsing structured data
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06F 16/3346 using probabilistic model
• CPC Class Codes: G06F 16/34 Browsing; Visualisation therefor
• CPC Class Codes: G06F 16/382 using citations hypermedia G06F16/94
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 16/95 Retrieval from the web
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 2216/11 Patent retrieval
• CPC Class Codes: Y10S 707/912 Applications of a database
• CPC Class Codes: Y10S 707/923 Intellectual property
• CPC Class Codes: Y10S 707/93 intellectual property analysis
• CPC Class Codes: Y10S 707/933 Citation analysis
• CPC Class Codes: Y10S 707/937 intellectual property searching

Abstract

In one embodiment the present invention provides a novel method for probabilistically quantifying a degree of relevance between two or more citationally or contextually related data objects, such as patent documents, non-patent documents, web pages, personal and corporate contacts information, product information, consumer behavior, technical or scientific information, address information, and the like. In another embodiment the present invention provides a novel method for visualizing and displaying relevance between two or more citationally or contextually related data objects. In another embodiment the present invention provides a novel search input/output interface that utilizes an iterative self-organizing mapping (“SOM”) technique to automatically generate a visual map of relevant patents and/or other related documents desired to be explored, searched or analyzed. In another embodiment the present invention provides a novel search input/output interface that displays and/or communicates search input criteria and corresponding search results in a way that facilitates intuitive understanding and visualization of the logical relationships between two or more related concepts being searched.

Claims

1. A method for measuring or quantifying a degree of relevance between two or more citationally or contextually related data objects, comprising the following steps:

measuring the number and type of relevance links between said two or more data objects of interest; and

applying a probability transform function to probabilistically determine or estimate an event probability that a particular defined relationship exists or will exist between said two or more data objects of interest, said transform function being formulated or determined through multi-variate regression analysis and wherein the existence or non-existence of the particular defined relationship is the dependent variable and wherein the number and type of relevance links is provided as one or more of the independent variables.

2. The method of claim 1 wherein said data objects comprise one or more of the following:

patent documents, non-patent documents, web pages, personal contact information, corporate contact information, product information, consumer behavior, technical or scientific information, or address information.

3. The method of claim 1 wherein said relevance links include one or more citational relationships between said two or more selected data objects occurring at either the first generation, the second generation or higher generations.

4. The method of claim 3 wherein said relevance links include one or more citational relationships between said two or more selected data objects occurring at the second generation and the third generation.

5. The method of claim 1 wherein said relevance links include one or more contextual relationships between said two or more selected data objects.

6. The method of claim 5 wherein said one or more contextual relationships includes word commonality, textual similarities, and/or word frequency similarity.

7. The method of claim 6 wherein the contextual similarity between said two or more data objects is measured by:

i) identifying a list of words used in each data object along with calculated word frequencies;

ii) multiplying each corresponding word frequency to obtain a frequency product for each word;

iii) dividing each frequency product by one-half the sum of the squares of each corresponding word frequency; and

iv) taking the sum total of the result for each word in each data object.

8. The method of claim 1 wherein the defined relationship comprises a first-generation citational relationship.

9. The method of claim 1 wherein the defined relationship comprises a second-generation or higher citational relationship.

10. The method of claim 1 wherein the defined relationship comprises classification in one or more of the same subject matter categories.

11. The method of claim 1 wherein the defined relationship comprises a hyperlink.

12. The method of claim 1 wherein the defined relationship comprises a first-generation citational relationship and wherein said transform function is formulated or determined through multi-variate probit regression analysis wherein the existence or non-existence of a first-generation citational relationship is the dependent variable and wherein the number and type of relevance links comprise one or more of the independent regression variables.

13. A method for visualizing and displaying degrees of relevance between two or more citationally or contextually related data objects comprising the followings steps:

selecting a relevant group of data objects;

determining relevance scores for each data object relative to at least one other data object;

selecting and/or assigning space coordinates to each data object;

optimizing an arbitrary or scaled distance between each data object and each other data object so to substantially minimize the distance between data objects having high relevance and substantially maximize the distance between data objects having low relevance.

14. The method of claim 13 wherein said relevance scores are determined by measuring the number and type of relevance links between said two or more data objects of interest;

and applying a probability transform function to probabilistically determine or estimate an event probability that a particular defined relationship exists or will exist between said two or more data objects of interest.

15. The method of claim 13 wherein said arbitrary or scaled distance is scaled such that data objects having a citational relationship at the first generation will, on average, have an arbitrary scaled distance equal to unity.

16. An improved search method and algorithm for locating patent documents and/or other related documents of interest comprising the following steps:

selecting a first group of patents representing the closest known references to a particular technology or search topic of interest;

performing relevance analysis on the first group to generate a second group of relevant patents, each having an associated relevance score to the first group;

reviewing the second group of relevant patents and selectively adding any desired additional relevant patents to the first group;

repeating the second and third steps until a desired list of most relevant patents is identified.