Cross-domain emotion analysis method based on emotion polarity reinforcing semantics
A technology of emotional polarity and sentiment analysis, applied in the field of text analysis, to achieve the effect of enhancing emotional semantics, reducing artificially set text features, and narrowing differences between fields
Active Publication Date: 2018-04-27
YUNNAN UNIV
View PDF9 Cites 7 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, even though these methods consider the bridge function of shared words between domains, they only use single and simple metrics such as co-occurrence and word frequency to select shared words.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0027] figure 1 The method flowchart provided for the embodiment of the present invention, such as figure 1 As shown, the method may include the following steps:
[0028] Step 101: Extract terms and vectorize:
[0029] In the implementation of the present invention, for example, English text can extract unigram and bigrams words as source domain terms through word segmentation, part-of-speech restoration, and removal of stop words W S and target domain terms W T ; Based on word2vec, the term W S with W T Vectorized, denoted as , .
[0030] Step 102: extracting a shared word set;
[0031] First, statistics W S ∩ W T The word frequency of each word in the source domain and the target domain; the selected word frequency is greater than the preset threshold β words, of which β is a positive integer ，β It can be preferred that 2≤β≤10;
[0032] Secondly, in the implementation of the present invention, according to the HowNet polar dictionary, N positive words a...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention belongs to the field of text analysis, and discloses a cross-domain emotion analysis method based on emotion polarity reinforcing semantics. The method comprises the steps of 1, extracting and vectorizing lexical items of emotion texts of a source domain and a target domain; 2, selecting the lexical items which have strong emotion and are consistent in semantics between the source domain and the target domain as inter-domain shared words; 3, extending the emotion texts respectively based on the emotion polarities of the shared words, retraining word vectors, and reinforcing the emotion semantics; 4, automatically extracting emotion features of the texts based on a convolution neural network, and training a classier to classify the emotion texts of the target domain. By meansof the cross-domain emotion analysis method based on the emotion polarity reinforcing semantics, the consistency of the emotion polarities of the shared words and the inner-domain emotion semantics and the influence of the emotion polarities on the extraction and classification of the emotion features are considered, and the method better conforms to actual characteristics and requirements of cross-domain emotion analysis.
Description
[0001] The invention belongs to the field of text analysis, relates to a cross-domain sentiment analysis method, and more specifically relates to a selection of shared words in an emotional text and a method for enhancing emotional semantics based on the shared words. Background technique [0002] Emotional texts refer to texts with subjective emotional tendencies. The analysis of the emotional tendency of the text is an important technical basis for applications such as public opinion monitoring, word-of-mouth analysis, and topic monitoring. Cross-domain sentiment analysis studies the technical issue of how to make full use of relevant source domain samples for analysis under the condition that sentiment has topic correlation and domain correlation, and the samples in the target domain are sparse. [0003] The key to solving cross-domain sentimental text analysis is to narrow the difference between the source domain and the target domain, transfer the knowledge of the source ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/30G06F17/27
CPCG06F16/35G06F40/216
Inventor 姬晨李维华王翔郭延哺段云浩
Owner YUNNAN UNIV