CN110852062B - Method for automatically measuring group external attitude and internal attitude by using speech information - Google Patents
Method for automatically measuring group external attitude and internal attitude by using speech information Download PDFInfo
- Publication number
- CN110852062B CN110852062B CN201910986512.2A CN201910986512A CN110852062B CN 110852062 B CN110852062 B CN 110852062B CN 201910986512 A CN201910986512 A CN 201910986512A CN 110852062 B CN110852062 B CN 110852062B
- Authority
- CN
- China
- Prior art keywords
- vocabulary
- concept
- attribute
- attitude
- statement
- 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.)
- Active
Links
Images
Landscapes
- Machine Translation (AREA)
Abstract
The invention discloses a method for automatically measuring group external attitude and internal hidden attitude by using speech information, which comprises the following steps of firstly, simulating concepts, examples and attributes in a self-evaluation report and an internal hidden association test by defining concept vocabularies, example vocabularies and attribute vocabularies; then according to given concept vocabulary, example vocabulary and attribute vocabulary, finding out statements related to the example vocabulary and statements related to the concept vocabulary in the corpus respectively to obtain two statement sets; and finally, measuring the external appearance and the internal appearance by calculating the emotion scores stated in the corresponding statement sets and taking the average emotion score of the corresponding statement sets as the quantitative representation of the internal appearance bias and the external appearance bias. The invention avoids the measurement of real testees, saves manpower and material resources, simultaneously, real corpora can better reflect the attitude of the group, and more objects can be researched.
Description
Technical Field
The invention relates to a group attitude bias measurement method, and belongs to a key research problem in the fields of social computing and social psychology. The method utilizes large-scale speech information of the population to automatically measure the exomorphism and the endomorphism of the population.
Background
Social psychology studies have shown that people often exhibit attitudes in public places that differ from the true insidious attitude of their mind. Attitudes are beliefs that have a tendency to concepts or objects. Since attitudes may be an intrinsic motivation for people's interest, opinion, or follow-up behavior, understanding attitude biases of groups is essential for opinion analysis, automatic recommendations, and user portrayal related research. In current approaches, languages are widely used to mine attitudes. By analyzing the emotion and the semantics of the words of people, the attitudes of people on a certain event, object, person or concept can be mined.
In the psychological research, self-assessment reports and implicit association tests are respectively used for measuring the human population external implicit attitude and the human population implicit attitude, but both methods need the tested personnel to actively cooperate to carry out the experiment, and the measurement can be carried out only by a small-scale population. Therefore, for both experiments, measuring the exo-and endo-attitudes of large-scale populations requires a lot of manpower and money. Therefore, it is necessary to develop an automated method for measuring exo-and endo-attitudes in large populations.
The self-evaluation report is an explicit attitude measuring method, which is usually performed in the form of survey, questionnaire, etc., and the questions answered by the subject can reflect the explicit attitude of the subject. The main advantage of self-assessment reporting is to allow participants to describe their own experiences, rather than being inferred by observing the participants. However, the self-assessment reporting method has two major disadvantages, which can result in a false report for the subject: (1) bias of social expectations: this means that the subject wishes to present themselves in a socially acceptable manner, for example against racial discrimination, even if this is not their true attitude; (2) unconsciousness: this means that the subjects are not aware of the true attitude in their mind. For example, homosexuals do not express their true sexual orientation explicitly in a self-assessment report because they are unaware of this.
The implicit association test is a method for measuring implicit state degree, which is most widely used in the field of social psychology, and measures the degree of closeness of automatic association between concept vocabularies and attribute vocabularies through a classification task realized by a computer program, so as to measure the implicit state degree of an individual. The implicit association test takes the reaction time as a measurement standard, and a general experimental step is to display an attribute vocabulary, so that a testee can distinguish and classify as soon as possible and click keys of a keyboard to react, and the intensity of implicit attitude is quantified through the magnitude of effect.
Specifically, implicit association testing is usually performed on a computer, and labels of categories are respectively displayed on the upper left corner and the upper right corner of a computer display, and words to be classified are displayed in the center of the display. Implicit association testing typically includes five components, each of which contains a classification task. The first part is mainly to classify the attribute vocabulary as fast as possible and to respond by clicking the keyboard button; the second part is mainly to classify the example vocabularies as soon as possible and to react by clicking the keyboard button; the third part is that all the words appearing in the first two parts are randomly displayed and classified after being mixed, and simultaneously, a button on a keyboard is clicked to react; the fourth part is similar to the second part, except that the example vocabulary is changed into the attribute vocabulary; the fifth part is similar to the third part and also exchanges the example vocabulary into the attribute vocabulary. In the above testing process, the reaction time of each key reaction and the wrong key alignment can be automatically recorded by the computer. And finally, calculating the difference between the reaction time mean value of the incompatible group and the reaction time mean value of the compatible group, thereby obtaining the intensity of the buried state degree of the subject.
In order to measure the bias automatically, researchers analyze the emotion and the semantics of words in human languages through natural language processing technology, and therefore understanding of the bias of people is achieved. Researchers find statements on wikipedia material with bias including framework bias and cognition bias, and these biases can be determined by common linguistic clues. In addition, word embedding is also used for capturing bias on linguistic data at present, and correlation test based on word embedding simulates correlation strength in a hidden association test through semantic distance between words so as to measure the bias. However, these methods do not distinguish between exo-and endo-states. If the appearance of the exo-and endo-attitudes in human speech are different, then the attitudes identified based on current methods may be confusing between exo-attitudes and endo-attitudes.
Disclosure of Invention
In order to be able to identify the exo-and endo-attitudes of a population separately. The invention provides a new strategy, which relates key factors (such as concepts, examples, attributes and the like) in self-evaluation reports and implicit association tests with language features and automatically realizes measurement of the explicit attitude and the implicit attitude of large-scale people by using a natural language processing technology. Compared with the traditional psychological method, the method avoids the measurement of a real 'testee', saves manpower and material resources, simultaneously, the real corpus can better reflect the attitude of a group, and more objects can be researched.
In order to solve the technical problem, the invention provides a method for automatically measuring the exomorphism and the endomorphism of a population by using speech information, which comprises the following steps:
step one, defining a concept vocabulary set, an example vocabulary set and an attribute vocabulary set by simulating concepts, examples and attributes in an explicit self-evaluation report and an implicit association test in psychology;
finding out statements containing the concept vocabularies, the attribute vocabularies and the attribute vocabularies in the social media corpus of the given group to form a statement set of the given group according to the given concept vocabulary set, the example vocabulary set and the attribute vocabulary set; finding a set of statement forming statements 1 containing the concept vocabulary and the attribute vocabulary in a set of statements of a given population, finding a set of statement forming statements 2 containing the example vocabulary and the attribute vocabulary in a set of statements of a given population;
and thirdly, calculating the emotion score of each statement in the statement set 1 and the statement set 2 respectively through an emotion analysis tool, taking the average emotion score of the statement set 1 as the quantitative representation of the bias of the external attitude degree, taking the average emotion score of the statement set 2 as the quantitative representation of the bias of the internal attitude degree, and realizing the measurement of the external attitude degree and the internal attitude degree of the population.
Furthermore, in the first step of the method of the present invention, the meaning of the concept vocabulary set is the vocabulary naming the concept and the synonyms thereof; the meaning of the example vocabulary set is to include the vocabulary that constitutes the given concept; the meaning of the set of attribute words is to include words that can express the meaning of the attribute of a given attitude.
In the second step, the statement comprises a main predicate object or a main system table structure, and the main predicate object or the main system table structure is determined by analyzing a sentence structure through a Stanford syntax analysis tool.
In step three, for the statement of Chinese, the emotion analysis tool is Baidu emotion analysis API; for statements in English, the sentiment analysis tool is a Stanford syntax analysis tool.
The invention relates to a method for automatically measuring the exomorphism and the endomorphism of a population by using speech information, which comprises the following specific contents of the second step and the third step:
given a pair of concepts C i And C j And a pair of attributes A p And A q Definition of Cw i And Cw j Are respectively concept C i And C j Concept vocabulary set of (Ew) i And Ew j Are respectively concept C i And C j Example vocabulary set of (Aw) p And Aw q Are respectively an attribute A p And A q The attribute vocabulary set of (2); given a set of statements S, classifying the sentences according to the concept vocabulary and the attribute vocabulary contained in the sentences in the set S to order
Representing four sets of statements, the sentences in these sets of statements simultaneously containing one statement from set Cw i /Cw j /Ew i /Ew j And a concept vocabulary or example vocabulary from the attribute set Aw p ∪Aw q Is a vocabulary of properties, i.e., S Cw_i The representation includes a word from the concept set Cw i A statement set consisting of the words in (1) and sentences of a word from the attribute word set; s. the Cw_j The representation includes a set Cw of concept words j A statement set consisting of the words in (1) and sentences of a word from the attribute word set; s. the Ew_i The representation consists of a set Ew of Fan Lici i A statement set consisting of the words in (1) and sentences of a word from the attribute word set; s Ew_j The representation consists of a set Ew of Fan Lici j And a statement set of sentences of words from the vocabulary of the attribute word set.
Set of statements S Cw_i ，S Cw_j ，S Ew_i ，S Ew_j For measuring about concept C, respectively i External attitude of (2) with respect to concept C j External attitude of (2) with respect to concept C i And about concept C j Inner hidden state ofDegree; explicit attitude deviation ExplicitBias (C) i ,C j ,A p ,A q ) And implicit State deviation ImplicititBias (C) i ,C j ,A p ,A q ) The calculation is carried out according to the following formula:
wherein p (S) denotes the statement set S Cw_i ，S Cw_j ，S Ew_i ，S Ew_j The sentiment score of one of the statements s, the sentiment score being set to +1,0, -1 represents the positive sentiment, the neutral sentiment and the negative sentiment, respectively.
Compared with the prior art, the invention has the beneficial effects that:
the method does not need active cooperation of the testee, and can be finished by the testee unconsciously; the sample size of the tested user is large; the tested objects are diversified; historical data can be analyzed, and different scenes and crowds can be distinguished easily.
Drawings
FIG. 1 is a block diagram of an automated method for measuring the extrinsic attitude and the intrinsic attitude of a population, respectively, using verbal information, in accordance with the present invention;
fig. 2 is an experimental result comparing each pair of concepts on each corpus in the embodiment of the present invention, wherein a positive number (negative number) indicates that the bias is more biased toward the right (left) concept on one column. For example, on the first column, insect is the left concept and Flower is the right concept.
FIG. 3 (a) is the evolutionary trend of the population in twitter with respect to the exomorphism and the endosymosis of "floral-worm";
FIG. 3 (b) is the evolution trend of the population in twitter with respect to the exo-and endo-attitudes of "instrumental-weapons";
FIG. 3 (c) is the evolution trend of the population in tweets with respect to the exo-crypto-exo-ontogeny of "white-black;
FIG. 3 (d) is the evolution trend of the population in twitter with respect to the exomorphism and the endo morphism of "Chinese-USA";
FIG. 4 (a) is the evolution trend of the population on microblog about the exo-and endo-attitudes of "flower-worms";
FIG. 4 (b) is the evolution trend of the crowd on the microblog about the explicit attitude and implicit attitude of 'instrument-weapon';
FIG. 4 (c) is the evolution trend of the population on microblogs with respect to the explicit and implicit attitudes of "white-black";
FIG. 4 (d) is the evolution trend of the population on microblog about the explicit attitude and implicit attitude of "China-USA";
FIG. 5 is a graph showing the stability of explicit and implicit attitudes on twitter and microblog corpora as a result of the method of the present invention.
Detailed Description
The invention will be further described with reference to the following figures and specific examples, which are not intended to limit the invention in any way.
The overall framework of the method for automatically measuring the exo-attitude and the endo-attitude of the population by using the speech information is shown in figure 1. Firstly, simulating concepts, examples and attributes in a self-evaluation report and a hidden association test to define concept vocabularies, example vocabularies and attribute vocabularies; then according to given concept vocabulary, example vocabulary and attribute vocabulary, finding out statements related to the example vocabulary and statements related to the concept vocabulary in the corpus respectively to obtain two statement sets; and finally, calculating the emotion scores of the statements in the corresponding statement sets, and taking the average emotion score of each statement set as the quantitative representation of the implicit attitude bias and the explicit attitude bias.
The invention relates to a method for automatically measuring the exomorphism and the cryptomorphism of a population by using speech information, which comprises the following specific contents:
(1) Defining the words involved in the method of the invention
In the invention, 3 types of vocabulary sets (a concept vocabulary set, an example vocabulary set and an attribute vocabulary set) are defined, and 3 key factors (concepts, examples and attributes) in a self-evaluation report and a hidden association test method are simulated.
Concept vocabulary set: including naming this notional vocabulary and its synonyms, etc. For example, the set of concept words for the concept "china" may be: { "China", "the people's republic of China", … … }.
Example vocabulary set: including words that may constitute a given concept, etc. For example, an example word set of the concept "china" may be: { "five-star red flag", "Chinese nation", "Beijing", "Maozedong", "Renminbi", … … }.
Attribute vocabulary set: including words that may express the meaning of a given attitude attribute, etc. For example, a "positive" set of attribute words may be: { "happy", … … }.
(2) Matching self-assessment reports and implicit associative tests with statements containing a particular vocabulary
(a) Self-assessment reports, exo-attitudes, and statements containing concept vocabulary + Attribute vocabulary
In the present invention, the measure of the self-assessment report on the exogenic bias is converted into a linguistic criterion: if a statement refers to a concept vocabulary relative to an attribute vocabulary, the emotion of the statement is treated as an explicit attitude to the concept. For example, the emotion of the proposition that "flowers are not beautiful" is the negative exomorphism to the concept of "flowers".
(b) Implicit association test, implicit attitude, and statement containing "example vocabulary + Attribute vocabulary
In the invention, the measuring standard of the implicit attitude in the implicit association test method is also converted into a corresponding language standard: if a sentence refers to a paradigm vocabulary with respect to a property vocabulary, the emotion of the sentence is considered to be implicit to the concept containing the paradigm. For example, the emotion of the phrase "rose is beautiful" is the forward implicit attitude to the concept "flower".
In addition, the present invention selects a statement that contains a principal predicate or a principal family structure and refers to both the concept/paradigm vocabulary and the sentence of the attribute vocabulary.
In particular, by means of the Stanford syntactic analysis tool (https://stanfordnlp.github.io/CoreNLP/) And analyzing the sentence structure to determine the structure of the main object or the main system table.
(3) And calculating the external appearance attitude and the internal hiding attitude, and measuring the external appearance attitude and the internal hiding attitude of each statement set.
Given a pair of concepts C i And C j (e.g., floral-worm) and a pair of attributes A p And A q Definition of Cw i And Cw j Are respectively concept C i And C j The concept vocabulary set of (E), ew i And Ew j Are respectively concept C i And C j Example vocabulary set of (Aw) p And Aw q Are respectively an attribute A p And A q The property vocabulary set of (1).
For the statement set S of a given group, the invention firstly classifies the sentences according to the concept words and the attribute words contained in the sentences so as to enable the sentences to be classified
Representing four sets of statements, the sentences in these sets simultaneously containing one statement from set Cw i /Cw j /Ew i /Ew j And a concept vocabulary or example vocabulary from the attribute set Aw p ∪Aw q Is a vocabulary of properties, i.e., S Cw_i The representation includes a set Cw of concept words i A statement set consisting of the words in (1) and sentences of a word from the attribute word set; s. the Cw_j The representation includes a set Cw of concept words j A statement set consisting of the words in (1) and sentences of a word from the attribute word set; s Ew_i The representation consists of a set Ew of Fan Lici i And a statement set consisting of sentences of words from the attribute word set; s Ew_j The representation consists of a set Ew of Fan Lici j And a word from the set of attribute wordsA set of statements made of sentences of (a). As explained above in the present invention, the set of statements S Cw_i ，S Cw_j ，S Ew_i ，S Ew_j For measuring about concept C, respectively i External attitude of (2) with respect to concept C j The exo-attitude of (2), with respect to concept C i And about concept C j Is hidden.
Specifically, the explicit attitude deviation and the implicit attitude deviation are calculated by the following formulas:
where p (S) represents the sentiment score of statement S, which may be set of statements S Cw_i ，S Cw_j ，S Ew_i ，S Ew_j Is specifically set out by S ∈ S Cw_j ，s∈S Cw_i ，s∈S Ew_j ，s∈S Ew_i To determine, the sentiment score is set to +1,0, with-1 representing positive, neutral and negative sentiments, respectively.
Calculating the emotion score:
the Chinese sentence obtains the emotion score through a Baidu emotion analysis API (http:// ai. Baidu. Com/tech/nlp/sensory _ classify);
english sentence passing Stanford tool (https://stanfordnlp.github.io/CoreNLP/) To obtain an emotion score.
Study materials:
by selecting 4 different corpora (Chinese Wikipedia, english Wikipedia, chinese microblog, english twitter) for experiment, firstly, the explicit attitude and the implicit attitude on the 4 corpora are evaluated, and then, the evolution of the explicit attitude and the implicit attitude on the Chinese microblog and English twitter corpora is analyzed.
1. Static evaluation of exo-and endo-aspects
4 different corpora are selected, relating to different languages and media, and are respectively tested. Table 1 describes the statistical results of different corpora. Four pairs of concept pairs which have proved to be prejudiced by groups of people in psychological experiments are selected for experiments, including: insect vs. flowers (instrument vs. flower), weapon vs. musical instruments (Weapon vs. Instrument), black vs. white (Afro-American vs. Euro-American), and Chinese vs. USA (China vs. American).
Table 1: statistical results of different predictions
| Media type | English | Chinese character |
| Official text | Wikipedia: 99M | Wikipedia: 7M |
| Informal text | Twitter: 100M | Microblog: 77M |
Fig. 2 depicts experimental results on selected corpora and concept pairs, and from these results there are 4 main findings:
(1) The general differences are as follows: there were significant differences in the implicit and explicit attitude deviations, and these differences were consistent with classical psychological experimental reports: flowers are more active than insects; musical instruments are more aggressive than weapons; the white name is more recent than the black name associated with positive attributes.
(2) In terms of concept: compared with the non-social concept (flower-worm, musical instrument-weapon), the difference between the implicit attitude deviation and the explicit attitude deviation is more obvious on the social concept (black-white, china-usa).
(3) In terms of scenes: for social concept pairs, implicit and explicit biases in social media are closer than in wikipedia.
(4) In terms of language (culture): on both english corpuses, the explicit and implicit attitudes with respect to "black-white" have opposite polarities (which is consistent with the report of implicit association test), but this phenomenon does not appear on chinese corpuses. Similarly, there is relative polarity for explicit and implicit attitudes in "china-usa", but not in english corpus.
These observations, above, are consistent with classical psychological studies. Further, these results are also consistent with the assumptions in classical psychology about implicit cognition: the difference between implicit and explicit attitudes may be explained by a bias in social expectations.
2. Dynamic evaluation of exo-and endo-aspects
Another advantage of the present invention over traditional psychology methods is the ability to dynamically assess the evolution of exo-attitudes and endo-attitudes. In the experiment, the corpus data was divided by month, and a total of 36 months of data was included. Fig. 3 and 4 depict the evolutionary trends in twitter and microblog corpora with respect to the explicit and implicit attitudes of the four pairs of concept pairs, respectively. Fig. 5 is an experimental result on the stability of both endo-attitude and exo-attitude. Stability is the standard deviation of attitude over time.
While the present invention has been described with reference to the accompanying drawings, the present invention is not limited to the above-described embodiments, which are illustrative only and not restrictive, and various modifications which do not depart from the spirit of the present invention and which are intended to be covered by the claims of the present invention may be made by those skilled in the art.
Claims (3)
1. A method for automatically measuring the exo-attitudes and endo-attitudes of a population by using speech information, which comprises the following steps:
step one, defining a concept vocabulary set, an example vocabulary set and an attribute vocabulary set by simulating concepts, examples and attributes in an explicit self-evaluation report and an implicit association test in psychology;
the meaning of the concept vocabulary set is the vocabulary and synonyms thereof including naming the concept;
the meaning of the example vocabulary set is to include the vocabulary that constitutes the given concept;
the meaning of the attribute vocabulary set is that the vocabulary set comprises words which can express the meaning of the attribute of the given attitude;
step two, according to a given concept vocabulary set, an example vocabulary set and an attribute vocabulary set, finding out statements containing the concept vocabulary, the attribute vocabulary and the attribute vocabulary in the social media corpus of a given group to form a statement set of the given group; finding a set of statement forming statements 1 containing the concept vocabulary and the attribute vocabulary in a set of statements of a given population, finding a set of statement forming statements 2 containing the example vocabulary and the attribute vocabulary in a set of statements of a given population; the specific contents are as follows:
given a pair of concepts C i And C j And a pair of attributes A p And A q Definition of Cw i And Cw j Are respectively concept C i And C j The concept vocabulary set of (E), ew i And Ew j Are respectively concept C i And C j Example vocabulary set of (Aw) q And Aw q Are respectively an attribute A p And A q The attribute vocabulary set of (2);
given a set of statements S, classifying the sentences according to the concept vocabulary and the attribute vocabulary contained in the sentences in the set S to order
Representing four sets of statements, the sentences in these sets of statements simultaneously containing one from the setCw i /Cw j /Ew i /Ew j And a concept vocabulary or example vocabulary from the attribute set Aw p ∪Aw q The vocabulary of attributes of (a), i.e.,
S Cw_i the representation includes a word from the concept set Cw i A statement set consisting of the words in (1) and sentences of a word from the attribute word set;
S Cw_j the representation includes a set Cw of concept words j A statement set consisting of the words in (1) and sentences of a word from the attribute word set;
S Ew_i the representation consists of a set Ew of Fan Lici i A statement set consisting of the words in (1) and sentences of a word from the attribute word set;
S Ew_j the representation is composed of a set Ew from Fan Lici j And a statement set consisting of sentences of words from the attribute word set;
thirdly, calculating the emotion score of each statement in the statement set 1 and the statement set 2 through an emotion analysis tool, taking the average emotion score of the statement set 1 as the quantitative representation of the bias of the external attitude degree, taking the average emotion score of the statement set 2 as the quantitative representation of the bias of the internal attitude degree, and realizing the measurement of the external attitude degree and the internal attitude degree of the group; the specific contents are as follows:
set of statements S Cw_i ，S Cw_j ，S Ew_i ，S Ew_j For measuring about concept C, respectively i External attitude of (2) with respect to concept C j External attitude of (2) with respect to concept C i And about concept C j The degree of implicit attitude of (c);
explicit attitude deviation ExplicitBias (C) i ，C j ，A p ，A q ) And implicit State deviation ImplicititBias (C) i ，C j ，A p ，A q ) The calculation is carried out according to the following formula:
wherein p (S) represents a set of statements S Cw_i ，S Cw_j ，S Ew_i ，S Ew_j To say s, the sentiment score is set to +1,0, -1 represents positive, neutral and negative sentiments, respectively.
2. The method according to claim 1, wherein in step two, the statement comprises a predicate element or a predicate table structure, and the predicate element or the predicate table structure is determined by analyzing a sentence structure with a Stanford syntax analysis tool.
3. The method for automatically measuring the exoattitudes and the endo attitudes of a population by using speech information as claimed in claim 1, wherein in step three, for the statement in Chinese, the emotion analysis tool is Baidu emotion analysis API; for statements in English, the sentiment analysis tool is a Stanford syntax analysis tool.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910986512.2A CN110852062B (en) | 2019-10-17 | 2019-10-17 | Method for automatically measuring group external attitude and internal attitude by using speech information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910986512.2A CN110852062B (en) | 2019-10-17 | 2019-10-17 | Method for automatically measuring group external attitude and internal attitude by using speech information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110852062A CN110852062A (en) | 2020-02-28 |
| CN110852062B true CN110852062B (en) | 2023-03-14 |
Family
ID=69597566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910986512.2A Active CN110852062B (en) | 2019-10-17 | 2019-10-17 | Method for automatically measuring group external attitude and internal attitude by using speech information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110852062B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112836486B (en) * | 2020-12-09 | 2022-06-03 | 天津大学 | Group hidden-in-place analysis method based on word vectors and Bert |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104850741A (en) * | 2015-05-06 | 2015-08-19 | 浙江师范大学 | Child-friendly implicit attitude test method and device |
| CN106951472A (en) * | 2017-03-06 | 2017-07-14 | 华侨大学 | A kind of multiple sensibility classification method of network text |
| CN109697284A (en) * | 2018-11-01 | 2019-04-30 | 天津大学 | A method of disclosing connoting emotions in Chinese corpus |
| CN110232343A (en) * | 2019-06-04 | 2019-09-13 | 重庆第二师范学院 | Children personalized behavioral statistics analysis system and method based on latent variable model |
-
2019
- 2019-10-17 CN CN201910986512.2A patent/CN110852062B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104850741A (en) * | 2015-05-06 | 2015-08-19 | 浙江师范大学 | Child-friendly implicit attitude test method and device |
| CN106951472A (en) * | 2017-03-06 | 2017-07-14 | 华侨大学 | A kind of multiple sensibility classification method of network text |
| CN109697284A (en) * | 2018-11-01 | 2019-04-30 | 天津大学 | A method of disclosing connoting emotions in Chinese corpus |
| CN110232343A (en) * | 2019-06-04 | 2019-09-13 | 重庆第二师范学院 | Children personalized behavioral statistics analysis system and method based on latent variable model |
Non-Patent Citations (1)
| Title |
|---|
| Measuring Bidirectional Subjective Strength of Online Social Relationship by Synthetizing the Interactive Language Features and Social Balance;Baixiang Xue等;《Springer》;20190207;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110852062A (en) | 2020-02-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Castillo et al. | How to “QuantCrit:” Practices and questions for education data researchers and users | |
| Ribeiro et al. | Decent work in Brazil: Context, conceptualization, and assessment | |
| Kunneman et al. | Signaling sarcasm: From hyperbole to hashtag | |
| Gigerenzer et al. | Surrogate science: The idol of a universal method for scientific inference | |
| Zumbo | Three generations of DIF analyses: Considering where it has been, where it is now, and where it is going | |
| Yu et al. | Compatibility between text mining and qualitative research in the perspectives of grounded theory, content analysis, and reliability. | |
| Boeve-de Pauw et al. | A cross-cultural study of environmental values and their effect on the environmental behavior of children | |
| Seih et al. | Development and examination of the linguistic category model in a computerized text analysis method | |
| Gottfried et al. | Friends and family: A literature review on how high school social groups influence advanced math and science coursetaking | |
| Cooper | The oppositional framing of bloggers | |
| Vander Putten et al. | Comparing results from constant comparative and computer software methods: A reflection about qualitative data analysis | |
| Ames et al. | Using SAS PROC MCMC for item response theory models | |
| Papoutoglou et al. | Linking personality traits and interpersonal skills to gamification awards | |
| CN110852062B (en) | Method for automatically measuring group external attitude and internal attitude by using speech information | |
| Gadermann et al. | Examining validity evidence of self-report measures using differential item functioning | |
| Wang et al. | What surprises, interests and confuses researchers? A frame-based analysis of knowledge emotion markers in research articles | |
| Calderon Carvajal et al. | Reviewing the structure of Kolb’s learning style inventory from factor analysis and Thurstonian Item Response Theory (IRT) model approaches | |
| McAuliff et al. | Taking the next steps: Promoting open science and expanding diversity in Law and Human Behavior. | |
| ALAVI et al. | Academic discipline DIF in an English language proficiency test | |
| Nikadon et al. | BERTAgent: The development of a novel tool to quantify agency in textual data | |
| Georgakalou et al. | Evaluating Leaders’ Strategic Thinking and Entrepreneurial Characteristics Using Semantic Analysis | |
| Çelik et al. | Investigation of conservatory students’ perceptions of value in terms of different variables | |
| Friedman | Data and visual displays in the Journal of Ecology 1996–2016 | |
| Izza et al. | UTILIZING TWITTER DATA TO UNDERSTAND GLOBAL HALAL INDUSTRY TRENDS AND DEVELOPMENTS IN THE DIGITAL ERA. | |
| Lin et al. | The House-Tree-Person test is not valid for the prediction of mental health: An empirical study using deep neural networks |
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 |