CN113486657A - Emotion-reason pair extraction system based on knowledge assistance - Google Patents

Emotion-reason pair extraction system based on knowledge assistance Download PDF

Info

Publication number
CN113486657A
CN113486657A CN202110841439.7A CN202110841439A CN113486657A CN 113486657 A CN113486657 A CN 113486657A CN 202110841439 A CN202110841439 A CN 202110841439A CN 113486657 A CN113486657 A CN 113486657A
Authority
CN
China
Prior art keywords
emotion
clause
clauses
reason
knowledge
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110841439.7A
Other languages
Chinese (zh)
Other versions
CN113486657B (en
Inventor
刘德喜
赵凤园
万常选
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhaoyang Health Guangzhou Technology Co ltd
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN202110841439.7A priority Critical patent/CN113486657B/en
Publication of CN113486657A publication Critical patent/CN113486657A/en
Application granted granted Critical
Publication of CN113486657B publication Critical patent/CN113486657B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/284—Lexical analysis, e.g. tokenisation or collocates
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/205—Parsing
    • G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/237—Lexical tools
    • G06F40/242—Dictionaries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/253—Grammatical analysis; Style critique
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Biomedical Technology (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Machine Translation (AREA)

Abstract

The invention discloses an emotion-reason pair extraction system based on knowledge assistance, and belongs to the technical field of natural language processing emotion reason prediction. A sentiment-reason pair extraction system based on knowledge assistance comprises the following three steps: emotion clause extraction, reason clause extraction and emotion-reason pairing, wherein the three steps are represented by knowledge-assisted word coding. The system is added with an external knowledge auxiliary system for learning, which is beneficial to the extraction of emotion-reason pairs to a certain extent, and effectively solves the problem of insufficient judgment on causal relationship between sentences in the current learning model.

Description

Emotion-reason pair extraction system based on knowledge assistance
Technical Field
The invention belongs to the technical field of natural language processing emotion reason prediction, and particularly relates to an emotion-reason pair extraction system based on knowledge assistance.
Background
In the existing method, candidate clauses or candidate clause pairs are mostly expressed in a vector mode and then sent into a deep learning model to predict whether causal relationships exist among the clauses.
The current method has three disadvantages. First, for a text containing a large number of clauses, the emotion-cause pairs of candidates share a pair, and therefore, the recognition efficiency is low, and the text is not suitable for a long text containing a large number of clauses. Although the ECPE-2D model employs some limiting rules, there are still more candidate emotion-cause pairs. Secondly, although the current model can improve the emotion-reason pair recognition effect through the interaction of the candidate emotion clause and the candidate reason clause, the interference condition exists, and is directly reflected on the experimental result: compared with a model for independently extracting emotion clauses on the same ECPE data set, the emotion clause extraction effect is generally and obviously reduced when emotion-reason combined extraction is adopted; and under the condition of manually giving the emotion clauses, the reason clause extraction effect is obviously superior to the reason clause extraction effect when the emotion-reason pair is adopted for extraction. Thirdly, regarding text emotion analysis, more artificial knowledge can help to improve the extraction effect at present, the reason for triggering emotion is mostly events, the main body of emotion is entities such as people, organizations and the like, as shown in fig. 1, and the characteristics are not fully utilized by a model.
Disclosure of Invention
1. Technical problem to be solved
The invention aims to provide an emotion-reason pair extraction system based on knowledge assistance, which aims to solve the problems in the background technology:
after word-level encoding of a text, the machine may not be able to identify the problem of the emotion words in the clause more accurately.
2. Technical scheme
The system introduces a manually constructed language and psychological characteristic knowledge base and the like for auxiliary coding, strengthens the identification of emotional words and psychological characteristics, and improves the extraction effect of emotional clauses. Meanwhile, part-of-speech labels including entity identification are added, information such as characters and events in the text is captured, and richer features are provided for extracting emotion and emotion reasons. Third, emotion and emotional cause are often co-occurring, meaning that if a clause is identified as an emotional clause with a greater probability, there is also at least one reason clause in its context with a greater probability. Therefore, the external knowledge auxiliary system learning is added, and the emotion-reason pair extraction is facilitated to a certain extent.
A system for emotion-cause pair extraction based on knowledge assistance, comprising the steps of:
s1, extracting emotion clauses;
s2, extracting reason clauses;
s3, emotion-reason pairing;
S1-S3 are all represented by knowledge-aided word encoding.
Preferably, the knowledge-assisted word encoding representation consists of 3 parts: BERT-based semantic encoding, LIWC linguistic psychology knowledge base-based part-of-speech encoding, and NLPIR-based part-of-speech encoding, wherein,
based on the semantic coding of BERT, each word w in the clause is coded by a BERT BASE modeljEncoding to obtain 768-dimensional word vector representation
Figure BDA0003178980130000021
Based on word class coding of a language psychological characteristic knowledge base of LIWC, an SC-LIWC dictionary (comprising 71 categories of human sensory words, emotional history words, cognitive history words, social history words and the like) constructed by the golden orchid and the like is adopted to carry out word w in clauses according to one-hotjEncoding is performed to obtain a 71-dimensional vector representation
Figure BDA0003178980130000022
Based on the part-of-speech coding of NLPIR, keeping { personal name nr, place name ns, other nouns n, adjective a, adverb d, verb v, pronoun rr, adverb,Other pronouns r and other parts of speech other } are 9 parts of speech, and the word w in the clause is treated according to one-hotjIs encoded to obtain a 9-dimensional vector representation
Figure BDA0003178980130000023
Preferably, the knowledge-assisted word encoding means that the semantic encoding of the BERT of the current word, the part-of-speech encoding of the LIWC linguistic-psychological characteristic knowledge base, and the part-of-speech encoding of the NLPIR are concatenated, and the calculation formula is as follows:
Figure BDA0003178980130000031
wherein xjA vector representation representing a word.
Preferably, the S1 uses a Bi-LSTM model with two layers of word layer and clause layer to encode and represent the clause, and performs binary bounding prediction, that is, if the model already identifies the emotion clause in the text d, there is an emotion clause
Figure BDA0003178980130000032
The clauses of (2) are set with the recognition results corresponding to the clauses
Figure BDA0003178980130000033
1, and the recognition results of the other clauses are 0; if the recognition results of all clauses in the text d are 0, namely
Figure BDA0003178980130000034
Then will be
Figure BDA0003178980130000035
Maximum first two clause recognition results
Figure BDA0003178980130000036
Set to 1 and the recognition results of the remaining clauses to 0.
Preferably, the calculation formula of S1 is as follows:
Figure BDA0003178980130000037
Figure BDA0003178980130000038
Figure BDA0003178980130000039
wherein the content of the first and second substances,
Figure BDA00031789801300000310
is a clause ciIs used to indicate that the emotion is encoded,
Figure BDA00031789801300000311
is a clause ciIs indicative of the context of the user,
Figure BDA00031789801300000312
is the predicted probability of an emotional clause.
Preferably, the S2 uses a Bi-LSTM model with two layers of a word layer and a clause layer to encode and represent the clause, and concatenates the clause with the emotion clause encoded representation, and then performs binary bounding prediction, that is, if the model identifies a reason clause in the text d, there is a reason clause
Figure BDA00031789801300000313
The clauses of (2) are set with the recognition results corresponding to the clauses
Figure BDA00031789801300000314
1, and the recognition results of the other clauses are 0; if the recognition results of all clauses in the text d are 0, namely
Figure BDA00031789801300000315
Then will be
Figure BDA00031789801300000316
Maximum first two clause recognition results
Figure BDA00031789801300000317
Set to 1 and the recognition results of the remaining clauses to 0.
Preferably, the calculation formula of S2 is as follows:
Figure BDA0003178980130000041
Figure BDA0003178980130000042
Figure BDA0003178980130000043
wherein the content of the first and second substances,
Figure BDA0003178980130000044
is a clause ciIs used to indicate that the emotion is encoded,
Figure BDA0003178980130000045
is a clause ciIs indicative of the context of the user,
Figure BDA0003178980130000046
the predicted probability of the reason clause.
Preferably, in S3, the clause is encoded and represented by a Bi-LSTM model with two layers, i.e., a term layer and a clause layer, and prediction probabilities and distance information of emotion clauses and reason clauses are added and then sent to a logistic regression model for prediction.
Preferably, the calculation formula of S3 is as follows:
Figure BDA0003178980130000047
Figure BDA0003178980130000048
preferably, the distance information is calculated as follows: setting emotion clauses
Figure BDA0003178980130000049
And reason clause
Figure BDA00031789801300000410
Is d relative toi,jJ-i, and the maximum number of clauses in all texts does not exceed M sentences. Initializing a 2M x 50 dimensional array with each row conforming to a normal distribution function, then vdRepresents the (d) th in the arrayi,j+ M) rows, which are applied to the test dataset by continuous training of the dataset to obtain a final representation of each relative position.
3. Advantageous effects
Compared with the prior art, the invention has the advantages that:
(1) emotion-cause pair extraction (ECPE) with greater accuracy
The evaluation results of ECPE-KA on the emotion-cause pair extraction task EPCE are shown in Table 1. As can be seen from Table 1, the ECPE-KA is significantly higher than the ECPE-2Steps and the RANKCP model in the accuracy P and F1 values, and is respectively higher than the F1 values by 18.84% and 4.59%; although ECPE-KA is slightly lower than the TDGC model in the accuracy rate P, the recall rate R is obviously improved, so that the F1 value of ECPE-KA is better than that of the TDGC model, and the improvement of the recall rate also indicates that the model obtains more correct clause pairs.
Compared with the most advanced model ECPE-2D at present, the ECPE-KA (F1-0.6914) achieves better effect than the ECPE-2D model (F1-0.6889) on the ECPE task, the accuracy P is improved by 0.85%, and the recall rate only has the defect of 0.19%.
TABLE 1 results of the experimental evaluation
Figure BDA0003178980130000051
(2) Reduced number of pairs of candidate affective reasons
The binarization process adopted by ECPE-KA ensures that each text has at least one candidate emotion-reason pair to be sent into calculation, and the number of the candidate emotion-reason pairs in the three submodels of ECPE-2step is less than 1, which means that the model has serious defects in emotion clause extraction or reason clause extraction, so that the pairing number is sharply reduced, but a plurality of possible correct candidate emotion-reason pairs are inevitably lost.
Therefore, the ECPE-KA model not only ensures that the emotion clauses and reason clauses which are as accurate as possible are extracted, but also reduces the number of investigation candidate emotion-reason pairs and improves the identification efficiency.
(3) ECPE-KA is more accurate in emotion reason extraction (ECE)
In a classical ECE task, emotion clauses are manually marked, and an ECPE-KA model does not require manual marking of the emotion clauses in a test set.
Table 2 shows that the ECPE-KA model is only lower in accuracy than CANN and PAE-DGL, superior in recall to all reference models, and finally differs from the best result (CANN) by only 2% at 1 without sentiment clauses annotated to the test data set. This shows that the method proposed herein can overcome the application limitation problem of manual emotion clause annotation on ECE mission, and certainly there is room for improvement.
This document compares the CANN-E model, which is a label that removes the emotion clauses in the data set from the CANN model that performs better under test. As is clear from Table 2, the performance of the CANN-E model after the emotion labels are removed is reduced linearly, and compared with the CANN model, the performance is reduced by 47.74% in the value of 1. And the ECPE-KA also achieves 0.7083 in the case of no emotional clause label, and the value of 1 is improved by 86.54 percent compared with the CANN-E.
TABLE 2 evaluation of emotional cause extraction tasks
Figure BDA0003178980130000061
Figure BDA0003178980130000071
Drawings
FIG. 1 is an example of emotional cause text;
FIG. 2 is a block diagram of a knowledge-aided emotion-cause pair extraction system;
fig. 3 shows a structure of a knowledge-aided clause representation.
Detailed Description
An emotion-reason pair extraction system (ECPE-KA) based on knowledge assistance is provided by combining an external artificial knowledge LIWC (language-mental feature) knowledge base and an NLPIR (nlPIR) part-of-speech analysis platform. The ECPE-KA system structure is shown in FIG. 2.
Example 1: knowledge-assisted clause representation
The knowledge-aided clause representation structure is shown in fig. 3. Given a text d ═ c containing | d | clauses1,c2,…,c|d|}, each clause
Figure BDA0003178980130000072
Respectively contain | ci| words. Each word wjIs represented by the code xjThe method comprises three parts, namely semantic coding based on BERT, part of speech coding based on an LIWC language psychological characteristic knowledge base and part of speech coding based on NLPIR.
(1) The ECPE-KA model first adopts the BERT BASE model to carry out the operation on each word w in the clausejEncoding to obtain 768-dimensional word vector representation
Figure BDA0003178980130000073
(2) Since the text is directed to a chinese dataset, the SC-LIWC dictionary constructed by golden blue et al is employed. The SC-LIWC dictionary includes 71 categories such as human sensory part of speech, emotional history part of speech, cognitive history part of speech, and social history part of speech. One-hot pair of words w in clauses is adopted in the textjEncoding is performed to obtain a 71-dimensional vector representation
Figure BDA0003178980130000074
(3) Because only entities such as names, pronouns of names and the like need to be identified in a focused mode to assist extraction of emotion clauses, and in order to avoid dimension sparseness caused by excessive part-of-speech types, the ECPE-KA model only adopts one type and part of two types of parts-of-speech, three types of parts-of-speech which are described in detail are removed, 8 types of parts-of-speech including the names nr, the names ns of places, other nouns n, adjective a, adverb, verb v, pronouns rr and other pronouns r are reserved finally after screening, and the rest one type of parts-speech is combined into other parts-of-speech other uniformly. One-hot is adopted in the text to the word w in the clausejIs encoded to obtain a 9-dimensional vector representation
Figure BDA0003178980130000081
In the ECPE-KA model, a word w in a candidate clausejIs encoded by
Figure BDA0003178980130000082
And
Figure BDA0003178980130000083
expressed as:
Figure BDA0003178980130000084
example 2: sentiment clause extraction
The extraction of emotion clauses adopts a two-layer Bi-LSTM model of a word layer and a clause layer:
(1) word layer Bi-LSTM
One will contain | ciClause of | words
Figure BDA0003178980130000085
Coded representation of
Figure BDA0003178980130000086
As input, send into Bi-LSTM model to get clause ciMiddle j (th) wordHidden layer representation of language
Figure BDA0003178980130000087
Obtaining a clause c by adopting a self-attention mechanism for each wordiCoded representation of
Figure BDA0003178980130000088
Figure BDA0003178980130000089
Where F represents a Bi-LSTM network using the self-attention mechanism.
(2) Clause layer Bi-LSTM
The purpose of the clause layer Bi-LSTM is to capture semantic dependencies between clauses. For text containing | d | clauses, d ═ c1,c2,…,c2,…,c|d|Coding each clause
Figure BDA00031789801300000810
Sending the hidden state into a Bi-LSTM model to obtain the hidden state of the Bi-LSTM, namely a clause ciIs represented by the context of
Figure BDA00031789801300000811
Figure BDA00031789801300000812
Finally will be
Figure BDA00031789801300000813
Enter softmax function to get clause ciProbability of being an emotional clause
Figure BDA00031789801300000814
Figure BDA0003178980130000091
Considering that at least one emotion clause exists in the text and most of the text contains at most two emotion clauses, in the binary bounding phase, the ECPE-KA model considers two cases: if the model has recognized an emotion clause in the text d, then
Figure BDA0003178980130000092
The clauses of (2) are set with the recognition results corresponding to the clauses
Figure BDA0003178980130000093
1, and the recognition results of the other clauses are 0; if the recognition results of all clauses in the text d are 0, namely
Figure BDA0003178980130000094
Then will be
Figure BDA0003178980130000095
Maximum first two clause recognition results
Figure BDA0003178980130000096
Set to 1 and the recognition results of the remaining clauses to 0.
Thus, a candidate emotion clause set in d is obtained
Figure BDA0003178980130000097
Example 3: reason clause extraction
The extraction of the reason clauses also adopts a Bi-LSTM with a word layer and a clause layer, wherein the coding structure of the clauses (the coding of the Bi-LSTM with the word layer) is the same as the clause coding structure in the emotion clause extraction stage.
Clause ciCoded representation of
Figure BDA0003178980130000098
And the emotion prediction probability value obtained in the first stage
Figure BDA0003178980130000099
Make a spliceTo obtain a clause ciCoded representation of
Figure BDA00031789801300000910
To capture context information, a vector representation of | d | clauses in text d is presented herein
Figure BDA00031789801300000911
As an input to the Bi-LSTM model, the hidden state of Bi-LSTM, i.e., clause c, is obtainediIs represented by the context of
Figure BDA00031789801300000912
Figure BDA00031789801300000913
Finally will be
Figure BDA00031789801300000914
Sending into softmax function to obtain clause ciIs predicted to have a probability value
Figure BDA00031789801300000915
Figure BDA00031789801300000916
Similar to binarization adopted by emotion clause extraction, considering that most texts contain at most two reason clauses, binarization of the reason clause extraction result is also divided into two cases: if the model has identified a reason clause in the text d, then there is
Figure BDA00031789801300000917
The clauses of (2) are set with the recognition results corresponding to the clauses
Figure BDA00031789801300000918
1, and the recognition results of the other clauses are 0; if all clauses in the text d are recognizedAre all 0, i.e.
Figure BDA00031789801300000919
Then will be
Figure BDA00031789801300000920
Maximum first two clause recognition results
Figure BDA00031789801300000921
Set to 1 and the recognition results of the remaining clauses to 0.
Thus, a candidate reason clause set in d is obtained
Figure BDA0003178980130000101
Example 4: emotion-reason pairing
For the set of emotion clauses in document d
Figure BDA0003178980130000102
And reason clause set
Figure BDA0003178980130000103
Performing a cartesian product to obtain all possible pairing results:
Figure BDA0003178980130000104
obtaining candidate emotion clauses by adopting text representation method in section 1
Figure BDA0003178980130000105
Coded representation of
Figure BDA0003178980130000106
And candidate reason clause
Figure BDA0003178980130000107
Coded representation of
Figure BDA0003178980130000108
Distance v for joining two clauses simultaneouslydPrediction probability of candidate emotion clause
Figure BDA0003178980130000109
And predicted probability of candidate reason clause
Figure BDA00031789801300001010
As a feature, the five codes are spliced to obtain an input vector of the emotion-reason pair filtering model
Figure BDA00031789801300001011
Comprises the following steps:
Figure BDA00031789801300001012
distance feature vdThe calculation method of (c) is as follows: setting emotion clauses
Figure BDA00031789801300001013
And reason clause
Figure BDA00031789801300001014
Is d relative toi,jJ-i, and the maximum number of clauses in all texts does not exceed M sentences. Initializing a 2M x 50 dimensional array with each row conforming to a normal distribution function, then vdRepresents the (d) th in the arrayi,j+ M) rows, which are applied to the test dataset by continuous training of the dataset to obtain a final representation of each relative position.
Then inputting the vector
Figure BDA00031789801300001015
Sending the sentence into a Logistic regression (Logistic) model to detect whether the two clauses have a causal relationship, and filtering to obtain an emotion-reason pair set:
Figure BDA00031789801300001016
retention
Figure BDA00031789801300001017
As a final emotion-cause pair extraction result.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and the preferred embodiments of the present invention are described in the above embodiments and the description, and are not intended to limit the present invention. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (10)

1. A system for emotion-cause pair extraction based on knowledge assistance, comprising the steps of:
s1, extracting emotion clauses;
s2, extracting reason clauses;
s3, emotion-reason pairing;
S1-S3 are all represented by knowledge-aided word encoding.
2. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 1, wherein: the knowledge-assisted word encoding representation consists of 3 parts: BERT-based semantic encoding, LIWC linguistic psychology knowledge base-based part-of-speech encoding, and NLPIR-based part-of-speech encoding, wherein,
based on the semantic coding of BERT, each word w in the clause is coded by a BERT BASE modeljEncoding to obtain 768-dimensional word vector representation
Figure FDA0003178980120000011
Part of speech coding based on LIWC language psychological characteristic knowledge base, SC-LIWC dictionary (including human) constructed by golden orchid and the like71 categories such as sense word category, emotion history word category, cognitive history word category and social history word category) according to one-hot pair clausesjEncoding is performed to obtain a 71-dimensional vector representation
Figure FDA0003178980120000012
Based on the part-of-speech coding of NLPIR, 9 parts-of-speech (including a person name nr, a place name ns, other nouns n, an adjective a, an adverb d, a verb v, a person pronoun rr, other pronouns r and other parts-of-speech other) are reserved, and words w in the clause are subjected to one-hot codingjIs encoded to obtain a 9-dimensional vector representation
Figure FDA0003178980120000013
3. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 2, wherein: the knowledge-assisted word coding means that the semantic coding of BERT of the current word, the part of speech coding of an LIWC language psychological characteristic knowledge base and the part of speech coding of NLPIR are spliced, and the calculation formula is as follows:
Figure FDA0003178980120000021
wherein xjA vector representation representing a word.
4. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 1, wherein: s1 adopts a two-layer Bi-LSTM model of a word layer and a clause layer to encode and express the clause and carries out binary bounding prediction, namely if the model identifies the emotion clause in the text d, the emotion clause is known
Figure FDA0003178980120000022
The clauses of (2) are set with the recognition results corresponding to the clauses
Figure FDA0003178980120000023
1, and the recognition results of the other clauses are 0; if the recognition results of all clauses in the text d are 0, namely
Figure FDA0003178980120000024
Then will be
Figure FDA0003178980120000025
Maximum first two clause recognition results
Figure FDA0003178980120000026
Set to 1 and the recognition results of the remaining clauses to 0.
5. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 4, wherein: the calculation formula of S1 is as follows:
Figure FDA0003178980120000027
Figure FDA0003178980120000028
Figure FDA0003178980120000029
wherein the content of the first and second substances,
Figure FDA00031789801200000210
is a clause ciIs used to indicate that the emotion is encoded,
Figure FDA00031789801200000211
is a clause ciIs indicative of the context of the user,
Figure FDA00031789801200000212
is the predicted probability of an emotional clause.
6. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 1, wherein: s2 adopts a two-layer Bi-LSTM model of a word layer and a clause layer to encode and express the clause, the clause is spliced with the emotion clause encoding and expressing, and then binary bounding prediction is carried out, namely if the model identifies the reason clause in the text d, namely, the reason clause exists
Figure FDA00031789801200000213
The clauses of (2) are set with the recognition results corresponding to the clauses
Figure FDA00031789801200000214
1, and the recognition results of the other clauses are 0; if the recognition results of all clauses in the text d are 0, namely
Figure FDA00031789801200000215
Then will be
Figure FDA00031789801200000216
Maximum first two clause recognition results
Figure FDA00031789801200000217
Set to 1 and the recognition results of the remaining clauses to 0.
7. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 6, wherein: the calculation formula of S2 is as follows:
Figure FDA0003178980120000031
Figure FDA0003178980120000032
Figure FDA0003178980120000033
wherein the content of the first and second substances,
Figure FDA0003178980120000034
is a clause ciIs used to indicate that the emotion is encoded,
Figure FDA0003178980120000035
is a clause ciIs indicative of the context of the user,
Figure FDA0003178980120000036
the predicted probability of the reason clause.
8. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 1, wherein: and S3, coding and expressing the clauses by adopting a two-layer Bi-LSTM model of a word layer and a clause layer, adding the prediction probability and distance information of the emotion clauses and reason clauses, and then sending the prediction probability and distance information to a logistic regression model for prediction.
9. The system for emotion-reason pair extraction based on knowledge assistance as claimed in claim 8, wherein: the calculation formula of S3 is as follows:
Figure FDA0003178980120000037
Figure FDA0003178980120000038
10. according to the claimsSolving 8 the emotion-reason pair extraction system based on knowledge assistance, which is characterized in that: the distance information is calculated as follows: setting emotion clauses
Figure FDA0003178980120000039
And reason clause
Figure FDA00031789801200000310
Is d relative toi，jJ-i, and the maximum number of clauses in all texts does not exceed M sentences, initializing a 2M multiplied by 50 dimensional array with each row conforming to a normal distribution function, and then vdRepresents the (d) th in the arrayi，j+ M) rows, which are applied to the test dataset by continuous training of the dataset to obtain a final representation of each relative position.
CN202110841439.7A 2021-07-26 2021-07-26 Emotion-reason pair extraction system based on knowledge assistance Active CN113486657B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110841439.7A CN113486657B (en) 2021-07-26 2021-07-26 Emotion-reason pair extraction system based on knowledge assistance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110841439.7A CN113486657B (en) 2021-07-26 2021-07-26 Emotion-reason pair extraction system based on knowledge assistance

Publications (2)

Publication Number Publication Date
CN113486657A true CN113486657A (en) 2021-10-08
CN113486657B CN113486657B (en) 2023-01-17

Family

ID=77943576

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110841439.7A Active CN113486657B (en) 2021-07-26 2021-07-26 Emotion-reason pair extraction system based on knowledge assistance

Country Status (1)

Country Link
CN (1) CN113486657B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114676259A (en) * 2022-04-11 2022-06-28 哈尔滨工业大学 Conversation emotion recognition method based on causal perception interactive network

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110472047A (en) * 2019-07-15 2019-11-19 昆明理工大学 A kind of Chinese of multiple features fusion gets over news viewpoint sentence abstracting method
CN110781369A (en) * 2018-07-11 2020-02-11 天津大学 Emotional cause mining method based on dependency syntax and generalized causal network
CN111382565A (en) * 2020-03-09 2020-07-07 南京理工大学 Multi-label-based emotion-reason pair extraction method and system
CN111581396A (en) * 2020-05-06 2020-08-25 西安交通大学 Event graph construction system and method based on multi-dimensional feature fusion and dependency syntax
CN111859957A (en) * 2020-07-15 2020-10-30 中南民族大学 Method, device and equipment for extracting emotion reason clause labels and storage medium
CN111914556A (en) * 2020-06-19 2020-11-10 合肥工业大学 Emotion guiding method and system based on emotion semantic transfer map
CN112183064A (en) * 2020-10-22 2021-01-05 福州大学 Text emotion reason recognition system based on multi-task joint learning
US20210043222A1 (en) * 2019-08-06 2021-02-11 Honda Motor Co., Ltd. Information processing apparatus, information processing method, and storage medium
CN112364127A (en) * 2020-10-30 2021-02-12 重庆大学 Short document emotional cause pair extraction method, system and storage medium
CN112836515A (en) * 2019-11-05 2021-05-25 阿里巴巴集团控股有限公司 Text analysis method, recommendation device, electronic equipment and storage medium
WO2021135193A1 (en) * 2019-12-30 2021-07-08 华南理工大学 Visual object guidance-based social media short text named entity identification method

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110781369A (en) * 2018-07-11 2020-02-11 天津大学 Emotional cause mining method based on dependency syntax and generalized causal network
CN110472047A (en) * 2019-07-15 2019-11-19 昆明理工大学 A kind of Chinese of multiple features fusion gets over news viewpoint sentence abstracting method
US20210043222A1 (en) * 2019-08-06 2021-02-11 Honda Motor Co., Ltd. Information processing apparatus, information processing method, and storage medium
CN112836515A (en) * 2019-11-05 2021-05-25 阿里巴巴集团控股有限公司 Text analysis method, recommendation device, electronic equipment and storage medium
WO2021135193A1 (en) * 2019-12-30 2021-07-08 华南理工大学 Visual object guidance-based social media short text named entity identification method
CN111382565A (en) * 2020-03-09 2020-07-07 南京理工大学 Multi-label-based emotion-reason pair extraction method and system
CN111581396A (en) * 2020-05-06 2020-08-25 西安交通大学 Event graph construction system and method based on multi-dimensional feature fusion and dependency syntax
CN111914556A (en) * 2020-06-19 2020-11-10 合肥工业大学 Emotion guiding method and system based on emotion semantic transfer map
CN111859957A (en) * 2020-07-15 2020-10-30 中南民族大学 Method, device and equipment for extracting emotion reason clause labels and storage medium
CN112183064A (en) * 2020-10-22 2021-01-05 福州大学 Text emotion reason recognition system based on multi-task joint learning
CN112364127A (en) * 2020-10-30 2021-02-12 重庆大学 Short document emotional cause pair extraction method, system and storage medium

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
JIAXIN YU等: "A Mutually Auxiliary Multitask Model With Self-Distillation for Emotion-Cause Pair Extraction", 《IEEE》 *
刘德喜: "检索式自动问答研究综述", 《计算机学报》 *
覃俊: "子句级别的自注意力机制的情感原因抽取模型", 《中南民族大学学报(自然科学版)》 *
郑胜协: "基于深度学习的文本情绪原因发现方法的研究与实现", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 *
高清红: "跨语言文本情感原因发现研究", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114676259A (en) * 2022-04-11 2022-06-28 哈尔滨工业大学 Conversation emotion recognition method based on causal perception interactive network
CN114676259B (en) * 2022-04-11 2022-09-23 哈尔滨工业大学 Conversation emotion recognition method based on causal perception interactive network

Also Published As

Publication number Publication date
CN113486657B (en) 2023-01-17

Similar Documents

Publication Publication Date Title
US20190332670A1 (en) Method for Automatically Detecting Meaning and Measuring the Univocality of Text
Ghosh et al. Fracking sarcasm using neural network
CN107092596B (en) Text emotion analysis method based on attention CNNs and CCR
CN113094578B (en) Deep learning-based content recommendation method, device, equipment and storage medium
CN112631436B (en) Method and device for filtering sensitive words of input method
CN112818698B (en) Fine-grained user comment sentiment analysis method based on dual-channel model
CN112287197B (en) Method for detecting sarcasm of case-related microblog comments described by dynamic memory cases
CN112434161B (en) Aspect-level emotion analysis method adopting bidirectional long-short term memory network
CN115292461B (en) Man-machine interaction learning method and system based on voice recognition
CN111145903A (en) Method and device for acquiring vertigo inquiry text, electronic equipment and inquiry system
CN111339772B (en) Russian text emotion analysis method, electronic device and storage medium
CN114927177B (en) Medical entity identification method and system integrating Chinese medical field characteristics
CN114756675A (en) Text classification method, related equipment and readable storage medium
CN113486657B (en) Emotion-reason pair extraction system based on knowledge assistance
CN110929518A (en) Text sequence labeling algorithm using overlapping splitting rule
CN114416969A (en) LSTM-CNN online comment sentiment classification method and system based on background enhancement
Danner et al. Advancing mental health diagnostics: GPT-based method for depression detection
Ruposh et al. A computational approach of recognizing emotion from Bengali texts
Harris et al. Constructing a rhetorical figuration ontology
Hemmatirad et al. Detection of mental illness risk on social media through multi-level svms
CN111815426A (en) Data processing method and terminal related to financial investment and research
CN111639189A (en) Text graph construction method based on text content features
CN116432659A (en) False news detection method based on title text similarity and author implicit emotion
CN116070620A (en) Information processing method and system based on big data
CN115269846A (en) Text processing method and device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right

Effective date of registration: 20240311

Address after: Room 501, Room 401, No. 50 Zhongshan 7th Road, Jinhua Street, Liwan District, Guangzhou City, Guangdong Province, 510000 (for office use only)

Patentee after: Zhaoyang Health (Guangzhou) Technology Co.,Ltd.

Country or region after: China

Address before: 330013 mailuyuan campus, Jiangxi University of Finance and economics, 665 Yuping West Street, Changbei national economic and Technological Development Zone, Nanchang City, Jiangxi Province

Patentee before: Liu Dexi

Country or region before: China

TR01 Transfer of patent right