System and method for topic extraction and opinion mining

Patent metadata

Field	Value
Publication number	US-12086191-B2
Application number	US-201916419908-A
Country	US
Kind code	B2
Filing date	May 22, 2019
Priority date	Sep 28, 2009
Publication date	Sep 10, 2024
Grant date	Sep 10, 2024

Abstract

Methods, apparatus, and systems to determine a niche market of items or services, the first phase of which identifies a gap between demand and supply for a set of items. Session logs may be evaluated to compare transactions involving a specific item to those of a larger group of items. The resultant information identifies areas of high demand, but with low availability. The niche market information may be provided as direct merchandising items for sellers. In one example, the method generates niche market item web pages in specific categories. Additional methods, apparatus, and systems are disclosed.

First claim

What is claimed is:

1. A method comprising:
accessing target documents obtained from one or more client devices;
searching, by a topic extractor, the target documents to identify key phrases related to different topics;
ranking, by the topic extractor, the key phrases by applying a ranking algorithm that is computed as a function of Term Frequency-Inverse Document Frequency (TF-IDF) weights;
identifying key phrases having rankings that exceed a threshold, the identified key phrases indicating essential topics, the identifying being based on a key phrase extraction process that builds a model based on training documents and uses the model to predict a likelihood of each phrase in a target document;
assigning, using one or more hardware processors, a polarity impact to one or more words of the target documents, the polarity impact being a score identifying an impact of the one or more words on a given topic of the essential topics;
identifying one or more opinion trends for at least one essential topic based on the polarity impact of the one or more words of the target documents by identifying a number of words for each polarity type over time for the at least one essential topic; and
causing display of a report in a user interface of a further client device, the report comprising one or more graphical indicators of the one or more opinion trends for the at least one essential topic, the report further indicating the polarity impact of the one or more words.

2. The method of claim 1 , further comprising:
determining that a word of the one or more words is a polarity word based on matching the word to another word in a dictionary of polarity words,
wherein the assigning of the polarity impact to the one or more words of the target documents is based on the polarity word having a dominant impact on a topic.

3. The method of claim 1 , wherein the assigning of the polarity impact to the one or more words of the target documents is based on a sum of polarities method.

4. The method of claim 1 , wherein the assigning of the polarity impact to the one or more words of the target documents is based on a syntactic distance between the one or more words and a topic key phrase in a syntactic tree.

5. The method of claim 1 , wherein the identifying of the one or more opinion trends based on the polarity impact of the one or more words includes generating a ratio of positive feedback to negative feedback.

6. The method of claim 1 , wherein the identifying of the one or more opinion trends based on the polarity impact of the one or more words includes:
identifying, using the polarity impact of the one or more words of the target documents, a first opinion associated with a first item and a second opinion associated with a second item; and
comparing the first opinion and the second opinion,
wherein the report includes an indication of a result of the comparing of the first opinion and the second opinion.

7. The method of claim 1 , wherein the report includes one or more sentences included in the target documents, the polarity impact of the one or more words being indicated graphically within the one or more sentences.

8. The method of claim 7 , wherein the graphic indication of the polarity impact of the one or more words within the one or more sentences includes representing words with positive impact with a first color and words with a negative impact with a second color.

9. The method of claim 1 , wherein the causing display of the report in the user interface of the further client device includes generating a corresponding graph by plotting a number of words for each polarity impact and presenting the graph in the user interface.

10. A system comprising:
one or more hardware processors; and
a non-transitory computer-readable medium storing instructions that, when executed by the one or more hardware processors, cause the one or more hardware processors to perform operations comprising:
accessing target documents obtained from one or more client devices;
searching, by a topic extractor, the target documents to identify key phrases related to different topics;
ranking, by the topic extractor, the key phrases by applying a ranking algorithm that is computed as a function of Term Frequency-Inverse Document Frequency (TF-IDF) weights;
identifying key phrases having rankings that exceed a threshold, the identified key phrases indicating essential top topics, the identifying being based on a key phrase extraction process that builds a model based on training documents and uses the model to predict a likelihood of each phrase in a target document;
assigning, using one or more hardware processors, a polarity impact to one or more words of the target documents, the polarity impact being a score identifying an impact of the one or more words on a given topic of the essential topics;
identifying one or more opinion trends for at least one essential topic based on the polarity impact of the one or more words of the target documents by identifying a number of words for each polarity type over time for the at least one essential topic; and
causing display of a report in a user interface of a further client device, the report comprising one or more graphical indicators of the one or more opinion trends for the at least one essential topic, the report further indicating the polarity impact of the one or more words.

11. The system of claim 10 , wherein the operations further comprise:
determining that a word of the one or more words is a polarity word based on matching the word to another word in a dictionary of polarity words,
wherein the assigning of the polarity impact to the one or more words of the target documents is based on the polarity word having a dominant impact on a topic.

12. The system of claim 10 , wherein the assigning of the polarity impact to the one or more words of the target documents is based on a sum of polarities method.

13. The system of claim 10 , wherein the assigning of the polarity impact to the one or more words of the target documents is based on a syntactic distance between the one or more words and a topic key phrase in a syntactic tree.

14. The system of claim 10 , wherein the identifying of the one or more opinion trends based on the polarity impact of the one or more words includes generating a ratio of positive feedback to negative feedback.

15. The system of claim 10 , wherein the identifying of the one or more opinion trends based on the polarity impact of the one or more words includes:
identifying, using the polarity impact of the one or more words of the target documents, a first opinion associated with a first item and a second opinion associated with a second item; and
comparing the first opinion and the second opinion,
wherein the report includes an indication of a result of the comparing of the first opinion and the second opinion.

16. The system of claim 10 , wherein the report includes one or more sentences included in the target documents, the polarity impact of the one or more words being indicated graphically within the one or more sentences.

17. The system of claim 16 , wherein the graphic indication of the polarity impact of the one or more words within the one or more sentences includes representing words with positive impact with a first color and words with a negative impact with a second color.

18. The system of claim 10 , wherein the causing display of the report in the user interface of the further client device includes presenting a corresponding graph representing a number of words for each polari

Assignees

• Ebay Inc

Inventors

• Sundaresan Neelakantan
• Zhang Yongzheng
• Baudin Catherine
• Shen Dan
• Huang Shen

Classifications

• G06F40/30

  Semantic analysis · CPC title

• G06Q30/00

  Commerce · CPC title

• G06Q10/00 Primary

  Administration; Management · CPC title

• G06F16/5846

  using extracted text · CPC title

• G06F16/3344

  using natural language analysis · CPC title