Perspective data analysis and management

• US 10,318,566 B2
• Filed: 09/24/2014
• Issued: 06/11/2019
• Est. Priority Date: 09/24/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Clark, Adam T., Huebert, Jeffrey K., Payton, Aspen L., Petri, John E.
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Chavez, Rodrigo A
• Application Number:
  US14/495,088
• Publication Number:
  US 20160085745A1
• Time in Patent Office: 1,721 Days
• Field of Search: 704 9
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

A system and computer implemented method for managing perspective data is disclosed. The method may include collecting a first lot of perspective data for an item. The method may include introducing a variant feature to the item to constitute a modified item. The method may include collecting a second lot of perspective data for the modified item. The method may also include evaluating the first and second lots of perspective data to ascertain a sentiment fluctuation based on information relevant to the variant feature.

Claims

1. A method comprising:

identifying, by a computing device with a processor, a variant feature of an item having a first set of reviews, wherein the variant feature is a changed characteristic of the item, and wherein identifying the variant feature comprises;

parsing, by the computing device, the first set of reviews using a natural language processing technique configured to analyze semantic content and syntactic content,

determining, by the computing device, sentiment factors for two or more reviews from the first set of reviews,

evaluating, by the computing device, the first set of reviews to identify a change in sentiment factor between the two or more reviews, and

identifying, by the computing device, the variant feature based on the change in sentiment factor;

grouping, by the computing device, based on the variant feature, the first set of reviews into a first group and a second group, wherein grouping the first set of reviews comprises;

analyzing, by the computing device, using natural language processing, the first set of reviews to determine whether reviews included in the first set of reviews are more closely associated with a first variable element of the variant feature or a second variable element of the variant feature, wherein the determining includes using a frequency analysis to determine whether terms or phrases related to the first variable element or the second variable element are mentioned with greater frequency in the reviews,

sorting, by the computing device, into the first group, based on the analyzing, first reviews of the first set of reviews that are associated with the first variable element of the variant feature, and

sorting, by the computing device, into the second group, based on the analyzing, second reviews of the first set of reviews that are associated with the second variable element of the variant feature;

determining, by the computing device, a first set of relevancy scores for the first group and a second set of relevancy scores for the second group by calculating weights of parsed content from each review based on a triggering event linked to the variant feature and computing the relevancy score using the calculated weights, wherein a relevancy score is a numerical value indicating a relative significance of a review;

establishing, by the computing device, using at least one of the first set of relevancy scores and the second set of relevancy scores, a second set of reviews configured as a subset of the first set of reviews, wherein the second set of reviews is a subset of the first set of reviews that have a relevancy score greater than a threshold relevancy score, and wherein the second set of reviews is a subset of the first set of reviews that are relevant and significant with respect to a user; and

providing, by the computing device, a visual representation of the second set of reviews to a display device, wherein the display device displays the visual representation to the user.

2. The method of claim 1, wherein:

evaluating the first set of reviews to identify the change in sentiment factor comprises determining the change in the sentiment factor in response to a triggering event, and

identifying the variant feature based on the change in sentiment factor comprises selecting the triggering event as the variant feature.

3. The method of claim 1, wherein the change in the sentiment factor is an increase in the sentiment factor greater than a sentiment change threshold.

4. The method of claim 1, wherein the change in the sentiment factor is a decrease in the sentiment factor greater than a sentiment change threshold.

5. The method of claim 1, wherein the variant feature is configured to change with respect to a predetermined fixed criterion.

6. The method of claim 1, wherein identifying the variant feature further comprises:

determining, in response to parsing the first set of reviews, a set of shared characteristics of the first set of reviews; and

selecting, from the set of shared characteristics of the first set of reviews, a first shared characteristic as the variant feature.

7. The method of claim 6, wherein selecting the first shared characteristic as the variant feature includes determining, based on semantic information analyzed by the natural language processing technique, that the first shared characteristic has a pertinence value greater than a pertinence threshold.

8. The method of claim 1, wherein establishing, using at least one of the first set of relevancy scores and the second set of relevancy scores, the second set of reviews includes:

determining that the second set of relevancy scores of the second group is lower than the threshold relevancy score; and

filtering the first set of reviews to remove the second group.

9. The method of claim 8, further comprising:

in response to determining that the second set of relevancy scores of the second relevancy group is lower than the threshold relevancy score, tagging the second group with an indicator to represent a lesser relevance.

10. The method of claim 1, wherein the visual representation of the second set of reviews is at least one of:

a chart, a graph, and a timeline.