APPARATUS AND METHOD FOR EXTRACTING SEMANTIC TOPIC

• US 20150268930A1
• Filed: 06/03/2015
• Published: 09/24/2015
• Est. Priority Date: 12/06/2012
• Status: Active Grant

• Current Assignee: Korea University Research and Business Foundation
• Original Assignee: Korea University Research and Business Foundation
• Inventors: Lee, Sang Keun, Alam, Md Hijbul
• Granted Patent: US 10,423,723 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 7/24 Sorting, i.e. extracting data from one or more carriers, rearranging the data in numerical or other ordered sequence, and rerecording the sorted data on the original carrier or on a different carrier or set of carriers sorting methods in generalG06F7/36 takes precedence
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

In accordance with a first exemplary embodiment, there is provided a method for extracting semantic topics from document sets in which opinions about an object are described using an apparatus capable of calculating a probability distribution. The method include (a) extracting word distributions about sentiment global topics and sentiment local topics; (b) extracting a global topic distribution, a local topic distribution and sentiment distributions about the global and local topics from the document sets; (c) performing statistical inference about each of the distributions extracted in the step (a) and the step (b); (d) extracting a global or local topic and a sentiment relevant to the global or local topic from the distributions of the inference performed in the step (c); and (e) extracting a word from the word distributions about sentiment global topics or sentiment local topics on the basis of the topic and sentiment extracted in the step (d).

Claims

1. A method for extracting a semantic topic from one or more document sets in which opinions about an object are described using an apparatus capable of calculating a probability distribution, the method comprising:

(a) extracting a word distribution about sentiment global topics and a word distribution about sentiment local topics;

(b) extracting a global topic distribution, a sentiment distribution about the global topic, a local topic distribution, and a sentiment distribution about the local topic with respect to each document of the document sets;

(c) performing statistical inference about each of the distributions extracted in the step (a) and the step (b);

(d) extracting a global or local topic and a sentiment relevant to the global or local topic from the global topic distribution and the sentiment distribution about the global topic or the local topic distribution and the sentiment distribution about the local topic with respect to each word in each document of the document sets; and

(e) extracting a word from the word distribution about sentiment global topics or the word distribution about sentiment local topics on the basis of the topic and sentiment extracted in the step (d).

2. The method of claim 1,

wherein the global topic refers to an aspect, which is used to distinguish the object from another object among aspects of the object, and

the local topic refers to a sentiment-oriented ratable aspect of the object.

3. The method of claim 1,

wherein each probability distribution is a multinomial distribution or a categorical distribution.

4. The method of claim 1,

wherein each probability distribution is constructed on the basis of a Dirichlet prior probability.

5. The method of claim 1,

wherein the statistical inference uses a Gibbs sampling algorithm.

6. The method of claim 1,

wherein the step (b) includes:

shifting one or more sliding windows overlapped with each other in each document; and

extracting the local topic distribution and the sentiment distribution about the local topic on the basis of a word extracted from a sentence in the sliding window,

wherein a size of the sliding window is set for a word expressing the local topic and a word expressing the sentiment about the local topic to be extracted together.

7. An apparatus for extracting a semantic topic, comprising:

a document storage unit that stores one or more document sets in which opinions about an object are described; and

a topic extraction unit that extracts a topic including a sentiment-oriented ratable aspect of the object and a sentiment about the topic from the document sets stored in the document storage unit,

wherein the topic extraction unit extracts a word distribution about sentiment topics, extracts a topic distribution and a sentiment distribution with respect to each document of the document sets, and extracts a topic and a sentiment from each of the extracted distributions with respect to each word in each document of the document sets.

8. The apparatus for extracting a semantic topic of claim 7,

wherein each probability distribution is constructed on the basis of a Dirichlet prior probability.

9. The apparatus for extracting a semantic topic of claim 7,

wherein the topic extraction unit uses Gibbs sampling in statistical inference of each of the already extracted distributions.

10. The apparatus for extracting a semantic topic of claim 7,

wherein the word distribution about sentiment topics includes a word distribution about sentiment global topics and a word distribution about sentiment local topics,

the topic distribution includes a global topic distribution and a local topic distribution, and

the sentiment distribution includes a sentiment distribution about a global topic and a sentiment distribution about a local topic.

11. The apparatus for extracting a semantic topic of claim 10,

wherein the global topic refers to an aspect, which is used to distinguish the object from another object, among aspects of the object, and

the local topic refers to a sentiment-oriented ratable aspect of the object.