Text analysis devices, articles of manufacture, and text analysis methods

• US 8,452,767 B2
• Filed: 09/15/2006
• Issued: 05/28/2013
• Est. Priority Date: 09/15/2006
• Status: Active Grant

• Current Assignee: Battelle Memorial Institute
• Original Assignee: Battelle Memorial Institute
• Inventors: Turner, Alan E., Hetzler, Elizabeth G., Nakamura, Grant C.
• Primary Examiner(s): Beausoliel, Jr., Robert
• Assistant Examiner(s): Pham, Michael
• Application Number:
  US11/522,178
• Publication Number:
  US 20080069448A1
• Time in Patent Office: 2,447 Days
• Field of Search: 707/736, 704/9, 704/4
• US Class Current: 707/736
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates

Abstract

Text analysis devices, articles of manufacture, and text analysis methods are described according to some aspects. In one aspect, a text analysis device includes processing circuitry configured to analyze initial text to generate a measurement basis usable in analysis of subsequent text, wherein the measurement basis comprises a plurality of measurement features from the initial text, a plurality of dimension anchors from the initial text and a plurality of associations of the measurement features with the dimension anchors, and wherein the processing circuitry is configured to access a viewpoint indicative of a perspective of interest of a user with respect to the analysis of the subsequent text, and wherein the processing circuitry is configured to use the viewpoint to generate the measurement basis.

Claims

1. A text analysis device comprising:

processing circuitry configured to analyze initial text to generate a measurement basis usable in analysis of subsequent text;

wherein the measurement basis comprises a plurality of measurement features from the initial text, a plurality of dimension anchors from the initial text and a plurality of associations of the measurement features with the dimension anchors;

wherein the processing circuitry is configured to access a viewpoint indicative of a perspective of interest of a user with respect to the analysis of the subsequent text, and wherein the processing circuitry is configured to use the viewpoint to generate the measurement basis; and

wherein the associations of the measurement basis comprise a plurality of association values indicative of probabilities that the dimension anchors are present in a plurality of text items of the subsequent text as a result of the identification of respective ones of the measurement features in the text items during the analysis of the subsequent text.

2. The device of claim 1 wherein the viewpoint comprises a plurality of text patterns, and the processing circuitry is configured to search for the presence of the text patterns during the analysis of the initial text to identify the measurement features and the dimension anchors.

3. The device of claim 2 wherein at least one of the measurement features comprises one of the text patterns.

4. The device of claim 2 wherein at least one of the dimension anchors is one of the text patterns.

5. The device of claim 2 wherein the viewpoint is configured to bias the analysis of the initial text by the processing circuitry to select one of the text patterns of the viewpoint as one of the measurement features compared with another text pattern not present in the viewpoint.

6. The device of claim 1 wherein the processing circuitry is configured to analyze the initial text comprising identifying a plurality of initial features occurring in the initial text, selecting some of the initial features as the measurement features and the dimension anchors, and generating the associations after the selecting.

7. The device of claim 6 wherein the viewpoint identifies at least one of the initial features.

8. The device of claim 7 wherein the processing circuitry is configured to:

score individual ones of the initial features; and

alter a score of one of the initial features responsive to the one initial feature being present within the viewpoint.

9. The device of claim 8 wherein the processing circuitry is configured to select the one initial feature as one of the measurement features using the score of the one initial feature.

10. The device of claim 6 wherein the viewpoint comprises a text pattern configured to bias the selection, by the processing circuitry, of at least one of the some initial features as one of the measurement features and which corresponds to the text pattern of the viewpoint.

11. The device of claim 1 wherein the processing circuitry is configured to alter respective ones of the association values which correspond to one of the measurement features responsive to the one of the measurement features corresponding to a text pattern of the viewpoint.

12. The device of claim 11 wherein the association values are in a matrix, and the processing circuitry is configured to alter selected ones of the association values of a row of the matrix corresponding to the one of the measurement features.

13. The device of claim 1 wherein the processing circuitry is configured to alter one of the association values of a first of the measurement features and a first of the dimension anchors responsive to the first measurement feature and the first dimension anchor corresponding to at least one text pattern of the viewpoint.

14. The device of claim 1 wherein the processing circuitry is configured to alter one of the association values of a first of the measurement features and a first of the dimension anchors responsive to a semantic similarity of the first measurement feature and the first dimension anchor.

15. The device of claim 1 wherein the perspective of interest comprises an initial perspective of interest and the measurement basis comprises an initial measurement basis configured to steer the analysis of the subsequent text with respect to the initial perspective of interest, and wherein the processing circuitry is configured to access another viewpoint indicative of another perspective of interest with respect to the analysis of the subsequent text, and wherein the processing circuitry is configured to generate another measurement basis using the another viewpoint and which is usable to steer the analysis of the subsequent text with respect to the another perspective of interest.

28. The device of claim 6 further comprising generating merit scores for respective ones of the initial features, and the selecting the initial features as measurement features comprises selecting using the merit scores, and the selecting the initial features as the dimension anchors comprises selecting the measurement features having the highest merit scores.

29. The device of claim 1 wherein the measurement basis comprises a matrix which includes the association values.

16. A text analysis device comprising:

processing circuitry configured to access a plurality of text items to be analyzed and to access a measurement basis comprising a plurality of relationships of text patterns and dimension anchors, and wherein the dimension anchors correspond to different topics of text content corresponding to a viewpoint indicative of a perspective of interest of a user;

wherein the processing circuitry is configured to analyze the text items using the measurement basis to associate the text items with respect to one another according to the perspective of interest; and

wherein the relationships of the measurement basis comprise a plurality of association values indicative of probabilities that the dimension anchors are present in the text items as a result of the identification of respective ones of the text patterns in the text items during the analysis of the text items.

17. The device of claim 16 wherein the viewpoint comprises a plurality of text patterns corresponding to the perspective of interest.

23. The device of claim 16 wherein the text items to be analyzed comprise subsequent text, and the processing circuitry is configured to access the viewpoint which comprises one or more text patterns defined at least in part by the user and to search for the presence of the one or more text patterns during analysis of initial text to identify the dimension anchors of the measurement basis.

30. The device of claim 16 wherein the measurement basis comprises a matrix which includes the association values.

18. An article of manufacture comprising:

non-transitory media storing programming configured to cause processing circuitry to perform processing comprising;

identifying a plurality of initial features of initial text;

analyzing the initial features to identify a plurality of measurement features and a plurality of dimension anchors;

using the identified measurement features and dimension anchors, generating a measurement basis configured to be used in analysis of subsequent text;

accessing a viewpoint indicative of a perspective of interest with respect to the analysis of the subsequent text;

utilizing the viewpoint to perform the generation of the measurement basis; and

wherein the measurement basis comprises a plurality of association values indicative of probabilities that the dimension anchors are present in a plurality of text items of the subsequent text as a result of the identification of respective ones of the measurement features in the text items during the analysis of the subsequent text.

24. The device of claim 18 wherein the accessing the viewpoint comprises accessing one or more text patterns defined at least in part by a user and the identifying and the analyzing comprise searching the initial text for the presence of the one or more text patterns to identify the measurement features and the dimension anchors.

25. The device of claim 18 wherein the analyzing comprises analyzing to identify the measurement features and the dimension anchors individually comprising a selected one of the initial features.

26. The device of claim 18 wherein the accessing the viewpoint comprises accessing a plurality of text patterns which are indicative of the perspective of interest from a user.

19. A text analysis method comprising:

accessing initial text;

accessing a viewpoint indicative of a perspective of interest of a user with respect to analysis of subsequent text;

steering analysis of the initial text using the viewpoint to provide a measurement basis configured to be used in the analysis of the subsequent text with respect to the perspective of interest, wherein the measurement basis comprises a plurality of associations of a plurality of text patterns with a plurality of dimension anchors in accordance with the perspective of interest of the viewpoint, and

wherein the associations of the measurement basis comprise a plurality of association values indicative of probabilities that the dimension anchors are present in a plurality of text items of the subsequent text as a result of the identification of respective ones of the text patterns in the text items during the analysis of the subsequent text.

20. The method of claim 19 wherein the steering comprises adjusting at least one of the association values corresponding to one of the text patterns and wherein the one of the text patterns is defined in the viewpoint.

21. The method of claim 19 wherein the viewpoint includes at least one text pattern, and wherein the steering comprises biasing selection of the at least one text pattern as one of the dimension anchors of the measurement basis.

27. The device of claim 19 wherein the accessing the viewpoint comprises accessing one or more text patterns which are defined at least in part by the user and the steering comprises searching the initial text for the presence of the one or more text patterns defined by the user to identify the text patterns and the dimension anchors of the measurement basis.

31. The method of claim 19 wherein the measurement basis comprises a matrix which includes the association values.

22. A text analysis device comprising:

processing circuitry configured to analyze initial text to generate a measurement basis usable in analysis of subsequent text;

wherein the measurement basis comprises a plurality of measurement features from the initial text, a plurality of dimension anchors from the initial text and a plurality of associations of the measurement features with the dimension anchors;

wherein the processing circuitry is configured to access a viewpoint which comprises one or more text patterns defined at least in part by a user and which are indicative of a perspective of interest of the user with respect to the analysis of the subsequent text, and wherein the processing circuitry is configured to use the viewpoint to generate the measurement basis;

wherein the processing circuitry is configured to search for the presence of the one or more text patterns during the analysis of the initial text to identify the measurement features and the dimension anchors; and

wherein the associations of the measurement basis comprise a plurality of association values indicative of probabilities that the dimension anchors are present in a plurality of text items of the subsequent text as a result of the identification of respective ones of the measurement features in the text items during the analysis of the subsequent text.

32. The device of claim 22 wherein the measurement basis comprises a matrix which includes the association values.