Systems and methods for finding project-related information by clustering applications into related concept categories

• US 9,256,422 B2
• Filed: 08/05/2014
• Issued: 02/09/2016
• Est. Priority Date: 09/29/2011
• Status: Active Grant

• Current Assignee: Accenture Global Services Limited (Accenture PLC)
• Original Assignee: Accenture Global Services Limited (Accenture PLC)
• Inventors: Grechanik, Mark
• Primary Examiner(s): Wong, Don
• Assistant Examiner(s): Rivera, Anibal
• Application Number:
  US14/452,433
• Publication Number:
  US 20140344783A1
• Time in Patent Office: 553 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 40/194 Calculation of difference between files
• CPC Class Codes: G06F 8/70 Software maintenance or management
• CPC Class Codes: G06F 8/71 Version control security arrangements therefor G06F21/57; Configuration management

Abstract

A system, method, and computer-readable medium, is described that finds similarities among programming applications based on semantic anchors found within the source code of such applications. The semantic anchors may be API calls, such as Java'"'"'s package and class calls of the JDK. Latent Semantic Indexing may be used to process the application and semantic anchor data and automatically develop a similarity matrix that contains numbers representing the similarity of one program to another.

Claims

1. A system, comprising:

a computer including;

a non-transitory memory device storing instructions and a similarity matrix, the similarity matrix combining;

a first categorization matrix defining a similarity between computer applications in a plurality of computer applications according to a first categorization of computer program calls,

a second categorization matrix defining a similarity between the computer applications according to a second categorization of the computer program calls, and

wherein the computer program calls are defined hierarchically and the first categorization corresponds to a first level of a hierarchy and the second categorization corresponds to a second level of the hierarchy; and

a processor that executes the instructions, causing the computer to;

receive a selection of one of the plurality of computer applications, and

indicate at least one of the plurality of computer applications

using the stored similarity matrix and based on the selected

one of the plurality of computer applications.

2. The system of claim 1, wherein the first categorization matrix defines the similarity between computer applications as a numerical value based on computer program calls in source code of the computer applications.

3. The system of claim 1, wherein the first categorization matrix is generated from a term document matrix using singular value decomposition, the term document matrix including columns corresponding to the computer applications and rows corresponding to categories of the first categorization.

4. The system of claim 3, wherein elements of the term document matrix have values based on a number of computer program calls in a computer application corresponding to the column and in the category corresponding to the row, at least one of the computer program calls corresponding to one of the categories.

5. The system of claim 4, wherein the elements of the term document matrix have values weighted by

the number of computer program calls in the computer application corresponding to the column, and

a number of computer applications in the plurality of computer applications that contain computer program calls in the category corresponding to the row.

6. The system of claim 1, wherein the second categorization matrix is generated from a plurality of vectors corresponding to the computer applications in the plurality of computer applications using a vector space model, the plurality of vectors including elements corresponding to categories of the second categorization, the elements including values based on a number of computer program calls in source code and documentation for a computer application corresponding to a vector and in the category corresponding to an element, at least one of the computer program calls corresponding to one of the categories.

7. The system of claim 1, wherein the hierarchy is an inheritance hierarchy.

8. The system of claim 1, wherein the first level of the hierarchy includes Java packages and the second level of the hierarchy includes Java classes.

9. The system of claim 1, wherein the first level of the hierarchy includes C++ namespaces and the second level of the hierarchy includes C++ classes.

10. The system of claim 1, wherein the computer program calls are API calls.

11. The system of claim 1, wherein the computer generates the stored similarity matrix.

12. The system of claim 10, wherein the computer receives the plurality of computer applications from a computer application archive over a network.

13. A system for determining similar computer applications, comprising:

one or more memories to store a plurality of computer applications;

one or more processor to implement;

a metadata extractor configured to receive the plurality of computer applications from the memory and to generate tuples associating at least one of the plurality of computer applications with computer program calls, the computer program calls being defined hierarchically with a first categorization corresponding to a first level of a hierarchy and a second categorization corresponding to a second level of the hierarchy;

a term document matrix builder configured to receive the tuples

and to generate a first term document matrix according to the first categorization and a second term document matrix according to the second categorization;

a similarity matrix builder configured to receive the first term document matrix and the second term document matrix and to generate a similarity matrix using singular value decomposition of the

first term document matrix and the second term document matrix; and

a search engine configured to receive a selection of one of the plurality of computer applications and provide an indication of at least

one of the computer applications using the similarity matrix and based on the selected one of the plurality of computer applications.

14. The system of claim 13, wherein the tuples comprise:

an indication of one of the plurality of computer applications,

an indication of one of the computer program calls,

an indication of a category of the first categorization encompassing the one of the computer program calls, and

an indication of a category of the second categorization encompassing the one of the computer program calls.

15. The system of claim 13, wherein the metadata extractor generates the tuples from source code of computer applications in the plurality of computer applications.

16. The system of claim 13, wherein the first term document matrix includes columns corresponding to computer applications in the plurality of computer applications, rows corresponding to categories of the first categorization, and elements of the term document matrix have values based on a number of computer program calls in a computer application corresponding to a column and in a category corresponding to a row, at least one of the computer program calls corresponding to one of the categories.

17. The system of claim 13, wherein the hierarchy is an inheritance hierarchy.

18. The system of claim 13, wherein the first level of the hierarchy includes Java packages and the second level of the hierarchy includes Java classes.

19. A method comprising:

generating a similarity matrix using a computer server, comprising;

receiving a plurality of computer applications from a computer application database,

generating tuples associating at least one of the plurality of computer applications with computer program calls, categories of a first categorization, and categories of a second categorization, wherein the computer program calls are defined hierarchically, the first categorization corresponding to a first level of the hierarchy and the second categorization corresponding to a second level of the hierarchy,

generating, based on the tuples, a first term document matrix according to the first categorization and a second term document matrix according to the second categorization, and

generating the similarity matrix from the first term document matrix and the second term document matrix, elements of the similarity matrix indicating a similarity between computer applications

corresponding to rows of the similarity matrix and computer applications corresponding to columns of the similarity matrix; and

providing an indication of relevant computer applications using the computer server, comprising;

providing a web service for processing code search requests,

indicating first relevant computer applications based on a code

search request, and

indicating second relevant computer applications using the similarity matrix based on a selection of one of the first relevant computer applications.

20. The method of claim 19, wherein indicating the second relevant computer applications comprises:

determining computer program calls shared by one of the second relevant computer applications and the selected one of the first relevant computer applications, and

providing instructions to display the shared computer program calls.