Method and system for analysing sentiments
First Claim
1. A computer-implemented method for training a document collection to analyze sentiment comprising:
- a) receiving a text expression from a computerized device;
b) a processor computing a frequency matrix from the received text expression, the frequency matrix comprising at least one row and at least column vectors, each row representing at least one term in a vocabulary each column representing at least one document of the document collection;
c) the processor executing a term reduction of the terms to remove useless terms in the frequency matrix;
d) the processor enumerating a plurality of categories of the document collection, the categories comprising at least 3 sentiment classes and at least 8 gradations for each sentiment classes, the classes comprising positive, negative, and neutral sentiments;
e) the processor computing centroids of the enumerated categories; and
f) the processor computing a fuzzy polarity map by;
i) building a polarity histogram of all terms belonging to the inputted expression;
ii) computing skewness of the polarity histogram;
iii) categorizing the inputted expression in terms of the computed skewness.
1 Assignment
0 Petitions
Accused Products
Abstract
The present system and method for analyzing sentiment is based on fuzzy set theory and clustering to classify text as positive, negative, or objective. The method for training and testing a document collection to analyze sentiment comprises computing a frequency matrix comprising at least one row and at least column vectors, executing term reduction of the terms, enumerating categories, computing centroids of the enumerated categories and computing a fuzzy polarity map. The row vectors may correspond to terms and the column vectors may correspond to documents. The frequencies of terms in a document indicate the relevance of the document to a query.
10 Citations
16 Claims
-
1. A computer-implemented method for training a document collection to analyze sentiment comprising:
-
a) receiving a text expression from a computerized device; b) a processor computing a frequency matrix from the received text expression, the frequency matrix comprising at least one row and at least column vectors, each row representing at least one term in a vocabulary each column representing at least one document of the document collection; c) the processor executing a term reduction of the terms to remove useless terms in the frequency matrix; d) the processor enumerating a plurality of categories of the document collection, the categories comprising at least 3 sentiment classes and at least 8 gradations for each sentiment classes, the classes comprising positive, negative, and neutral sentiments; e) the processor computing centroids of the enumerated categories; and f) the processor computing a fuzzy polarity map by; i) building a polarity histogram of all terms belonging to the inputted expression; ii) computing skewness of the polarity histogram; iii) categorizing the inputted expression in terms of the computed skewness. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
Specification