Topical sentiments in electronically stored communications
First Claim
1. A computer implemented method for obtaining topical sentiments from a plurality of electronically stored communications comprising the steps of:
- determining with the assistance of a computer whether each communication in a plurality of communications is topical to a first predefined topic;
for each communications determined to be topical to the predefined topic, separating with the assistance of a computer the communication into one or more expressions;
for each expression, determining with the assistance of a computer if the expression is topical to a second predefined topic; and
for each expression that is determined to be topical to the second predefined topic, determining with the assistance of a computer a polarity of the expression.
5 Assignments
0 Petitions
Accused Products
Abstract
The present application presents methods for performing topical sentiment analysis on electronically stored communications employing fusion of polarity and topicality. The present application also provides methods for utilizing shallow NLP techniques to determine the polarity of an expression. The present application also provides a method for tuning a domain-specific polarity lexicon for use in the polarity determination. The present application also provides methods for computing a numeric metric of the aggregate opinion about some topic expressed in a set of expressions.
329 Citations
51 Claims
-
1. A computer implemented method for obtaining topical sentiments from a plurality of electronically stored communications comprising the steps of:
-
determining with the assistance of a computer whether each communication in a plurality of communications is topical to a first predefined topic;
for each communications determined to be topical to the predefined topic, separating with the assistance of a computer the communication into one or more expressions;
for each expression, determining with the assistance of a computer if the expression is topical to a second predefined topic; and
for each expression that is determined to be topical to the second predefined topic, determining with the assistance of a computer a polarity of the expression. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A computer implemented method for calculating, from a plurality of electronically stored expressions, an aggregate metric which estimates a frequency of positive and/or negative polar expressions contained in the expressions, comprising the steps of:
-
for each of a plurality of electronically stored expressions, determining with the assistance of a computer an opinion contained in the expressions based upon at least one of the presence, absence and strength of polarity associated with a predefined topic; and
calculating an aggregate metric from the determined opinions of the plurality of expressions. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29)
-
-
30. A computer implemented method for finding one or more polar expressions in an electronically stored communication, comprising the step of analyzing with the assistance of a computer the electronically stored communication for one or more polar expressions within the electronically stored communication, wherein the analyzing step includes the steps of:
-
providing a polarity lexicon of sentimental words and/or phrases associated with a topical domain;
utilizing with the assistance of a computer the polarity lexicon against words and/or phrases found in the expression; and
assigning with the assistance of a computer at least one word/phrase in the expression a polarity associated with a matching word/phrase in the polarity lexicon. - View Dependent Claims (31, 32, 33, 34, 35)
-
-
36. A computer implemented method for obtaining topical sentiments from a body of electronically stored communications comprising the steps of:
-
isolating with the assistance of a computer a first subset of the communications topical to a particular topic; and
isolating with the assistance of a computer a second subset of communications from the first subset of communications where the second subset of communications includes polar segments located in the first subset of communications. - View Dependent Claims (37, 38, 39, 40, 41, 42)
-
-
43. A computer implemented method for tuning a polarity lexicon for use in classifying polar expressions, comprising the steps of:
-
providing a polarity lexicon;
with the assistance of a computer implemented graphical user interface providing a user with candidate words for addition, subtraction or exclusion to the polarity lexicon; and
adding, subtracting or excluding each candidate word from the polarity lexicon according to input received by the graphical user interface. - View Dependent Claims (44, 45, 46, 47, 48)
-
-
49. A computer implemented method for obtaining topical sentiments from an electronically stored communication comprising the steps of:
-
determining with the assistance of a computer one or more topical expressions in the communication;
locating with the assistance of a computer one or more polar expressions in the communication; and
identifying an expression that is both a topical expression and a polar expression as containing a topical sentiment. - View Dependent Claims (50, 51)
-
Specification