CN109726938A - A kind of students ' thinking political affairs situation method for early warning based on deep learning - Google Patents
A kind of students ' thinking political affairs situation method for early warning based on deep learning Download PDFInfo
- Publication number
- CN109726938A CN109726938A CN201910078081.XA CN201910078081A CN109726938A CN 109726938 A CN109726938 A CN 109726938A CN 201910078081 A CN201910078081 A CN 201910078081A CN 109726938 A CN109726938 A CN 109726938A
- Authority
- CN
- China
- Prior art keywords
- passive
- vocabulary
- early warning
- students
- political affairs
- 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
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of students ' thinking political affairs situation method for early warning based on deep learning, library is expected especially by set the whole network news data training, the retrieval record and keyword in student's internet behavior in retrieval record with emotional problems tendency phase are accurately analyzed, has reached that data are complete and the novel purpose of data.Gather intelligent algorithm and big data analysis technology simultaneously, entire think of political affairs status analysis process does not need artificially subjectively to participate in the task of analysis, can counselor school work is carried out to student or thinks the early warning of political affairs situation, it is more objective compared to more traditional thinks of political affairs status analysis, more with comprehensive.
Description
Technical field
The present invention relates to computer internet field more particularly to it is a kind of using computer internet technology realize based on
The students ' thinking political affairs situation method for early warning of deep learning.
Background technique
College student initial contact society, leaves home environment, and the various thoughts that touch society influence, and is easy to happen unstable
Variation, this is the unstable factor that the unstable factor of school is also society.
The form of traditional questionnaire is also limited to for the think of political affairs status analysis of student at present, it can not be accurate and real-time
The think of political affairs situation of student is analyzed on ground, and questionnaire be easy to cause wrong report to conceal, can not actual response student's heart.It is negative by special messenger
Duty notices students psychology variation, then there is very big subjectivity, and low efficiency accuracy rate is low.
Thought condition investigation is also broken off relations with achievement at present, the automatic judging method be combineding with each other without the two.
Summary of the invention
In order to solve problem above, the present invention proposes a kind of students ' thinking political affairs situation method for early warning based on deep learning, with
By computer internet technology, realizes to the quantitative classification of students ' thinking political affairs situation, avoid conventional method subjectivity big, judge to imitate
The low technical problem of rate.
The present invention is realized using following technical scheme:
A kind of students ' thinking political affairs situation method for early warning based on deep learning, comprising the following steps:
Step S1, the acquisition and classification of initial data
Step S101 constructs positive core vocabulary, negative core vocabulary and network politics sensitivity vocabulary database;
Step S102 intercepts the news for containing negative core vocabulary in title using web crawlers technology, is formed
Newsletter archive data;Word segmentation processing is carried out to newsletter archive data, obtains news sample, part-of-speech tagging is carried out to news sample,
Remove non-passive or nonsense words, passive vocabulary dictionary SAD is generated, by network politics sensitive word and passive vocabulary dictionary SAD
Merge, forms early warning dictionary;
Step S103 obtains enough network playing by students retrieval record samples, carries out word segmentation processing and carry out the product of part of speech
Pole/passiveness mark, to not can determine that it is actively/vocabulary of passive part of speech gives up in advance as neutral vocabulary, obtain actively/it is passive
Two classified lexicons;
Step S2, the processing of network playing by students data to be evaluated
Step S201 obtains student to be evaluated online retrieval record sample interior for a period of time, forms several short texts,
Record sample is retrieved in online temporally to divide;
Step S202 is filtered positive web search record with early warning dictionary, wherein containing word in early warning dictionary
The web search record of remittance is changed to negative sort;
Online is retrieved record sample and temporally divided by step S203, counts each period passiveness search record quantity and accounts for
The period always searches for the ratio of record quantity, judges the search condition of this period, judges to be measured according to search condition
Raw state, it is passive for normal, in a restless state of mind, long-term passive or change suddenly;
Step S3, the acquisition of student achievement data to be evaluated
The achievement data for taking student to be evaluated each term judges whether that fruitful fluctuation is big, is decreased obviously, is long-term poor
Situation;
Step S4, the judgement of students ' thinking political affairs situation
If student performance fluctuation is big, achievement is decreased obviously or achievement is poor for a long time, and the long-term passive or thought of thought is prominent
It is passive so to become, then carries out thinking the serious early warning of political affairs situation and academic warning;
If the student performance is normal, but in a restless state of mind, long-term passive or change suddenly is passive, then carries out thinking the early warning of political affairs situation;
If student performance fluctuation is big, achievement is decreased obviously or achievement is poor for a long time, but thought is normal, then it is pre- to carry out school work
It is alert;
If the student performance is normal, but in a restless state of mind, long-term passive or change suddenly is passive, then carries out thinking the slight early warning of political affairs.
It is further preferred that
In the step S102, the news in half a year is intercepted.
In the step S102, the method for part-of-speech tagging is that sentiment dictionary is generated using word2vec, and each vocabulary is corresponding
The vector of one n dimension: N=(x1,x2,...,xn);Calculate the term vector of the vocabulary in news sample and the word of negative core vocabulary
The meaning of a word degree of correlation len of vector, calculation formula are as follows:
If the term vector degree of correlation of any vocabulary in the vocabulary and negative core vocabulary is greater than 0.5, which is received
Enter in passive vocabulary dictionary SAD, and retain the maximum value of the word-correlativity as weight, negative core vocabulary is also received
Enter in passive vocabulary dictionary SAD.
In the step S103, enough network playing by students retrieval record samples refer at least the 5000 of at least 100 students
A sample.
It further include, to the vocabulary by positive/passive mark, being trained using LSTM algorithm in the step S103,
Two positive/passive categorization modules after being trained judge student network retrieval record sample to be evaluated using two disaggregated models
Whether this is passive.
In the step S201, the online retrieval record sample that student to be evaluated is more than 3 months is obtained, and by week segmentation.
In the step S202, method that positive web search record is filtered with early warning dictionary are as follows: containing pre-
The web search record of any vocabulary is changed to negative sort in alert dictionary, disappears in this web search record containing unduplicated
The weighted value of pole vocabulary is added, and result is changed to actively classify less than 1.5 to this web search record, the passiveness
The weighted value of vocabulary is the maximum value of the meaning of a word degree of correlation of the term vector of the vocabulary and the term vector of negative core vocabulary.
In the step S203, the judgment method of state are as follows:
The period for taking total search to be recorded as 3 or more is computing object, calculates the ratio that passive searching times account for total degree
Example thinks this period search condition for passiveness if more than 0.4;
If the search condition of continuous 5 periods it is normal with it is passive between convert, the Student Ideology fluctuation;
If the search condition of continuous 4 periods is passiveness, the Student Ideology is passive for a long time;
If the passive searching times Zhan of subsequent time period is always secondary after continuous 3 search conditions more than period are normal
Number is greater than 0.6, then the Student Ideology becomes passive suddenly.
In the step S3, the judgment method of achievement are as follows:
Step S301 obtains the marks sequencing ratio of each term various courses of student, and section [0,1], 1 is ranking the
One;This term ranking ratio average p (0) and last term average value p (- 1) are calculated, and so on;
Step S302, judgement:
P (- 1) p (- 2) if it exists, and meet p (- 1)-p (- 2) > 0.3 and p (- 1)-p > 0.3, then it is judged as achievement wave
It is dynamic big;
(- 1) p if it exists, and meet p (- 1)-p > 0.4, then it is judged as that achievement is decreased obviously;
(- 2) p if it exists, and meet p, p (- 1), p (- 2) is respectively less than 0.2, then be judged as that achievement is poor for a long time.
The beneficial effects of the present invention are:
This programme gives a kind of more specifically judges student performance and thought condition with thought condition on merit
Method is expected that library is accurately analyzed in student's internet behavior in retrieval record by gathering the training of the whole network news data and is asked with emotion
The retrieval record and keyword of topic tendency phase, have reached that data are complete and the novel purpose of data.Gather intelligent algorithm simultaneously
With big data analysis technology, entire think of political affairs status analysis process does not need artificially subjectively to participate in the task of analysis, can
Counselor carries out school work to student or thinks the early warning of political affairs situation, more objective compared to more traditional think of political affairs status analysis, more has
It is comprehensive.
Detailed description of the invention
Fig. 1 is overall structure diagram of the invention;
Fig. 2 is that political affairs situation judgment method schematic diagram is thought in the present invention.
Specific embodiment
The application is described in further detail with reference to the accompanying drawing, it is necessary to it is indicated herein to be, implement in detail below
Mode is served only for that the application is further detailed, and should not be understood as the limitation to the application protection scope, the field
Technical staff can make some nonessential modifications and adaptations to the application according to above-mentioned application content.
Embodiment 1
A kind of students ' thinking political affairs situation method for early warning based on deep learning, such as Fig. 1-2, comprising the following steps:
Step S1, the acquisition and classification of initial data
Step S101 constructs positive core vocabulary, negative core vocabulary and network politics sensitivity vocabulary database, wherein
Positive core vocabulary such as " fine " " love " " help ", negative core vocabulary such as " suicide " " depression " " unsociable and eccentric ", network politics
Deeply grateful word is a kind of dictionary that major website all has, and can be directly acquired;
Step S102, using web crawlers technology, to the news in the nearly half a year for containing negative core vocabulary in title into
Row interception, forms newsletter archive data;The newsletter archive data that the present embodiment obtains, specific as follows:
Word segmentation processing is carried out to newsletter archive data, such as existing jieba module can be used that sentence is divided into word, is obtained
Obtain the news sample being made of a large amount of vocabulary;
Part-of-speech tagging is carried out to news sample, removes non-passive or nonsense words, generates passive vocabulary dictionary SAD, tool
Body step includes:
Sentiment dictionary, the vector of the corresponding n dimension of each vocabulary: N=(x are generated using word2vec1,x2,...,xn);
Calculate the meaning of a word degree of correlation len of the term vector of the vocabulary in news sample and the term vector of negative core vocabulary, calculation formula are as follows:
If the term vector degree of correlation of any vocabulary in the vocabulary and negative core vocabulary is greater than 0.5, which is received
Enter in passive vocabulary dictionary SAD, and retain the maximum value of the word-correlativity as weight, negative core vocabulary is also received
Enter in passive vocabulary dictionary SAD.
Finally, network politics sensitive word is merged with passive vocabulary dictionary SAD, early warning dictionary is formed,
Step S103, by the information management system of school, backstage obtains nearly trimestral at least 100 students at least
5000 online retrieval record samples carry out word segmentation processing to sample and carry out the positive/passive of part of speech to mark, and method can be used
Artificial mark, or the method for generating sentiment dictionary using word2vec, to not can determine that it is actively/the vocabulary conduct of passive part of speech
Neutral vocabulary is given up in advance, obtains two classified lexicons of actively/passiveness;
It, can be to two positive/passive classified lexicons of acquisition in order to further increase the treatment effeciency of subsequent samples classification
Deep learning, including LSTM, SVM, Naive Bayes, CNN scheduling algorithm are carried out, two disaggregated models of actively/passiveness are obtained.
The present embodiment is tested in practice, uses sample 80% as training sample, and 20% is used as test sample, meter
The accuracy of various algorithm models is calculated, specific data are as follows:
It can be seen that LSTM has a preferable accuracy to short text, the present embodiment is trained with LSTM algorithm, obtain actively/
Two passive disaggregated models.
Step S2, the processing of network playing by students data to be evaluated
Step S201 obtains the online retrieval record sample of student's at least three moon to be evaluated, and form is several short essays
This, obtains data method particularly includes: plug-in unit is installed in outer net expenses of surfing in Internet management client, sniff http address and plaintext transmission
Data, and from sniff to webpage in search web page title, the code of Chinese character and Chinese character is therefrom found, as internet searching
Behavioral data.It can also directly be obtained by the backstage of School Information Management System bases.
Step S202 is filtered positive web search record with early warning dictionary, wherein containing word in early warning dictionary
The web search record of remittance is changed to negative sort;Specifically: the web search record containing vocabulary any in early warning dictionary is changed to
Negative sort is added the weighted value containing unduplicated passive vocabulary in this web search record, and result is less than
1.5 are changed to actively classify to this web search record.
Online is retrieved record sample by all segmentations, total search is taken to be recorded as 3 or more weeks as calculating pair by step S203
As, total week of the search record less than 3 is not included in calculating scope, the ratio that passive searching times account for total degree is calculated, if more than
0.4 is thought this period search condition for passiveness, and on the contrary is normal;
If the search condition of continuous 5 periods it is normal with it is passive between convert, the Student Ideology fluctuation;
If the search condition of continuous 4 periods is passiveness, the Student Ideology is passive for a long time;
If the passive searching times Zhan of subsequent time period is always secondary after continuous 3 search conditions more than period are normal
Number is greater than 0.6, then the Student Ideology becomes passive suddenly.
Step S3, the acquisition of student achievement data to be evaluated
The achievement data for taking student to be evaluated each term judges whether that fruitful fluctuation is big, is decreased obviously, is long-term poor
Situation;Step includes:
Step S301 transfers the marks sequencing ratio of each term various courses of student to be evaluated from school's education administration system
Example, section [0,1], 1 is to rank the first;Calculate this term ranking ratio average p (0) and last term average value p (-
1), last term average value p (- 2), and so on;
Step S302, judgement:
P (- 1) p (- 2) if it exists, and meet p (- 1)-p (- 2) > 0.3 and p (- 1)-p > 0.3, then it is judged as achievement wave
It is dynamic big;
(- 1) p if it exists, and meet p (- 1)-p > 0.4, then it is judged as that achievement is decreased obviously;
(- 2) p if it exists, and meet p, p (- 1), p (- 2) is respectively less than 0.2, then be judged as that achievement is poor for a long time.
Step S4, the judgement of students ' thinking political affairs situation
If student performance fluctuation is big, achievement is decreased obviously or achievement is poor for a long time, and the long-term passive or thought of thought is prominent
It is passive so to become, then carries out thinking the serious early warning of political affairs situation and academic warning;
If the student performance is normal, but in a restless state of mind, long-term passive or change suddenly is passive, then carries out thinking the early warning of political affairs situation;
If student performance fluctuation is big, achievement is decreased obviously or achievement is poor for a long time, but thought is normal, then it is pre- to carry out school work
It is alert;
If the student performance is normal, but in a restless state of mind, long-term passive or change suddenly is passive, then carries out thinking the slight early warning of political affairs.
The basic principles, main features and advantages of the invention have been shown and described above.The technical staff of the industry should
Understand, the present invention is not limited to the above embodiments, and the above embodiments and description only describe originals of the invention
Reason, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes and improvements
It all fall within the protetion scope of the claimed invention.The claimed scope of the invention is by appended claims and its equivalent circle
It is fixed.
Claims (9)
1. a kind of students ' thinking political affairs situation method for early warning based on deep learning, which comprises the following steps:
Step S1, the acquisition and classification of initial data
Step S101 constructs positive core vocabulary, negative core vocabulary and network politics sensitivity vocabulary database;
Step S102 intercepts the news for containing negative core vocabulary in title using web crawlers technology, forms news
Text data;Word segmentation processing is carried out to newsletter archive data, obtains news sample, part-of-speech tagging is carried out to news sample, is generated
Passive vocabulary dictionary SAD merges network politics sensitive word with passive vocabulary dictionary SAD, forms early warning dictionary;
Step S103 obtains enough network playing by students retrieval record samples, carry out word segmentation processing and carry out part of speech it is positive/disappear
Pole marks note, to not can determine that it is actively/vocabulary of passive part of speech gives up in advance as neutral vocabulary, obtain actively/passive two
Classified lexicon;
Step S2, the processing of network playing by students data to be evaluated
Step S201 obtains student to be evaluated online retrieval record sample interior for a period of time, forms several short texts, will be upper
Net is retrieved record sample and is compared one by one with two classified lexicons of actively/passiveness, and actively/passiveness two class web search note is obtained
Record;
Step S202 is filtered positive web search record with early warning dictionary;
Step S203 temporally divides online retrieval record sample, when counting each period passiveness search record quantity and accounting for this
Between section always search for record quantity ratio, judge the search condition of this period, which judged according to search condition
State, it is passive for normal, in a restless state of mind, long-term passive or change suddenly;
Step S3, the acquisition of student achievement data to be evaluated
The achievement data for taking student to be evaluated each term judges whether that fruitful fluctuation is big, is decreased obviously, long-term poor feelings
Condition;
Step S4, the judgement of students ' thinking political affairs situation
If student performance fluctuation is big, achievement is decreased obviously or achievement is poor for a long time, and the long-term passive or thought of thought becomes suddenly
Passiveness then carries out thinking the serious early warning of political affairs situation and academic warning;
If the student performance is normal, but in a restless state of mind, long-term passive or change suddenly is passive, then carries out thinking the early warning of political affairs situation;
If student performance fluctuation is big, achievement is decreased obviously or achievement is poor for a long time, but thought is normal, then carries out academic warning;
If the student performance is normal, but in a restless state of mind, long-term passive or change suddenly is passive, then carries out thinking the slight early warning of political affairs.
2. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S102, the news in half a year is intercepted.
3. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
State in step S102, the method for part-of-speech tagging is that sentiment dictionary is generated using word2vec, the corresponding n dimension of each vocabulary to
Amount: N=(x1,x2,...,xn);Calculate the meaning of a word of the term vector of the vocabulary in news sample and the term vector of negative core vocabulary
Degree of correlation len, calculation formula are as follows:
If the term vector degree of correlation of any vocabulary in the vocabulary and negative core vocabulary is greater than 0.5, which is included in and is disappeared
In the lexicon dictionary SAD of pole, and retain the maximum value of the word-correlativity as weight, negative core vocabulary is also included in and is disappeared
In the lexicon dictionary SAD of pole.
4. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S103, enough network playing by students retrieval record samples refer at least 5000 samples of at least 100 students.
5. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S103 and further includes, to the vocabulary by positive/passive mark, be trained using LSTM algorithm, after being trained
Two positive/passive categorization modules, judge that student network retrieval to be evaluated records whether sample is to disappear using two disaggregated models
Pole.
6. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S2, obtains the online retrieval record sample that student to be evaluated is more than 3 months, and by week segmentation.
7. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S202, the method that positive web search record is filtered with early warning dictionary are as follows: contain in early warning dictionary
The web search record of one vocabulary is changed to negative sort, to the power for containing unduplicated passive vocabulary in this web search record
Weight values are added, and result is changed to actively classify less than 1.5 to this web search record, the weight of the passive vocabulary
Value is the maximum value of the meaning of a word degree of correlation of the term vector of the vocabulary and the term vector of negative core vocabulary.
8. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S203, the judgment method of state are as follows:
The period for taking total search to be recorded as 3 or more is computing object, calculates the ratio that passive searching times account for total degree, if
Think this period search condition for passiveness greater than 0.4;
If the search condition of continuous 5 periods it is normal with it is passive between convert, the Student Ideology fluctuation;
If the search condition of continuous 4 periods is passiveness, the Student Ideology is passive for a long time;
If it is big that the passive searching times of subsequent time period account for total degree after continuous 3 search conditions more than period are normal
In 0.6, then the Student Ideology becomes passive suddenly.
9. a kind of students ' thinking political affairs situation method for early warning based on deep learning according to claim 1, which is characterized in that institute
It states in step S3, the judgment method of achievement are as follows:
Step S301 obtains the marks sequencing ratio of each term various courses of student, and section [0,1], 1 is to rank the first;Meter
This term ranking ratio average p (0) and last term average value p (- 1) are calculated, and so on;
Step S302, judgement:
P (- 1) p (- 2) if it exists, and meet p (- 1)-p (- 2) > 0.3 and p (- 1)-p > 0.3, then it is judged as that achievement fluctuation is big;
(- 1) p if it exists, and meet p (- 1)-p > 0.4, then it is judged as that achievement is decreased obviously;
(- 2) p if it exists, and meet p, p (- 1), p (- 2) is respectively less than 0.2, then be judged as that achievement is poor for a long time.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910078081.XA CN109726938B (en) | 2019-01-28 | 2019-01-28 | Student thinking state early warning method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910078081.XA CN109726938B (en) | 2019-01-28 | 2019-01-28 | Student thinking state early warning method based on deep learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109726938A true CN109726938A (en) | 2019-05-07 |
CN109726938B CN109726938B (en) | 2023-09-12 |
Family
ID=66301180
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910078081.XA Active CN109726938B (en) | 2019-01-28 | 2019-01-28 | Student thinking state early warning method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109726938B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111680926A (en) * | 2020-06-10 | 2020-09-18 | 桂林理工大学 | Big data-based student thinking state early warning method |
CN114331775A (en) * | 2022-03-15 | 2022-04-12 | 杭州职业技术学院 | Method for realizing college three-child-care through big data and block chain technology |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130103623A1 (en) * | 2011-10-21 | 2013-04-25 | Educational Testing Service | Computer-Implemented Systems and Methods for Detection of Sentiment in Writing |
CN107239439A (en) * | 2017-04-19 | 2017-10-10 | 同济大学 | Public sentiment sentiment classification method based on word2vec |
CN107506349A (en) * | 2017-08-04 | 2017-12-22 | 卓智网络科技有限公司 | A kind of user's negative emotions Forecasting Methodology and system based on network log |
CN108268448A (en) * | 2018-01-24 | 2018-07-10 | 闽南师范大学 | A kind of the analysis of public opinion method, system and medium based on Internet news |
CN108363699A (en) * | 2018-03-21 | 2018-08-03 | 浙江大学城市学院 | A kind of netizen's school work mood analysis method based on Baidu's mhkc |
-
2019
- 2019-01-28 CN CN201910078081.XA patent/CN109726938B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130103623A1 (en) * | 2011-10-21 | 2013-04-25 | Educational Testing Service | Computer-Implemented Systems and Methods for Detection of Sentiment in Writing |
CN107239439A (en) * | 2017-04-19 | 2017-10-10 | 同济大学 | Public sentiment sentiment classification method based on word2vec |
CN107506349A (en) * | 2017-08-04 | 2017-12-22 | 卓智网络科技有限公司 | A kind of user's negative emotions Forecasting Methodology and system based on network log |
CN108268448A (en) * | 2018-01-24 | 2018-07-10 | 闽南师范大学 | A kind of the analysis of public opinion method, system and medium based on Internet news |
CN108363699A (en) * | 2018-03-21 | 2018-08-03 | 浙江大学城市学院 | A kind of netizen's school work mood analysis method based on Baidu's mhkc |
Non-Patent Citations (1)
Title |
---|
刘志兵: "高校网络舆情监控系统的实现", 《长沙大学学报》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111680926A (en) * | 2020-06-10 | 2020-09-18 | 桂林理工大学 | Big data-based student thinking state early warning method |
CN114331775A (en) * | 2022-03-15 | 2022-04-12 | 杭州职业技术学院 | Method for realizing college three-child-care through big data and block chain technology |
Also Published As
Publication number | Publication date |
---|---|
CN109726938B (en) | 2023-09-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zhang et al. | Characterizing online discussion using coarse discourse sequences | |
CN110516067A (en) | Public sentiment monitoring method, system and storage medium based on topic detection | |
CN111178586B (en) | Method for tracking, predicting and dredging network patriotic public opinion events | |
Sharma et al. | Nlp and machine learning techniques for detecting insulting comments on social networking platforms | |
CN105354216B (en) | A kind of Chinese microblog topic information processing method | |
CN103049435A (en) | Text fine granularity sentiment analysis method and text fine granularity sentiment analysis device | |
Pota et al. | A subword-based deep learning approach for sentiment analysis of political tweets | |
CN104572877A (en) | Detection method and detection system of game public opinion | |
Weller et al. | The rJokes dataset: a large scale humor collection | |
Kawintiranon et al. | PoliBERTweet: a pre-trained language model for analyzing political content on Twitter | |
CN110825868A (en) | Topic popularity based text pushing method, terminal device and storage medium | |
Krishnan et al. | KnowSum: knowledge inclusive approach for text summarization using semantic allignment | |
CN109726938A (en) | A kind of students ' thinking political affairs situation method for early warning based on deep learning | |
Wang et al. | Multi‐label emotion recognition of weblog sentence based on Bayesian networks | |
TW201640383A (en) | Internet events automatic collection and analysis method and system thereof | |
Liang et al. | Enhancing content marketing article detection with graph analysis | |
CN111598691B (en) | Method, system and device for evaluating default risk of credit/debt main body | |
CN113468868A (en) | NLP-based real-time network hotspot content analysis method | |
CN111859955A (en) | Public opinion data analysis model based on deep learning | |
Gaugaz et al. | Predicting the future impact of news events | |
Zhu et al. | YUN111@ Dravidian-CodeMix-FIRE2020: Sentiment Analysis of Dravidian Code Mixed Text. | |
CN114742281A (en) | Public event network public opinion popularity prediction method based on grey model | |
CN114443930A (en) | News public opinion intelligent monitoring and analyzing method, system and computer storage medium | |
Gritz et al. | On the impact of features and classifiers for measuring knowledge gain during web search-a case study | |
Riordan | Comparing semantic space models using child-directed speech |
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 | ||
CB02 | Change of applicant information | ||
CB02 | Change of applicant information |
Address after: 1218, 12th floor, building 8, East District, yard 9, Linglong Road, Haidian District, Beijing 100089 Applicant after: BEIJING TAOHUADAO INFORMATION TECHNOLOGY Co.,Ltd. Address before: Room 1503, Yanshan Hotel, No. 38 Guancun Avenue, Haidian District, Beijing Applicant before: BEIJING TAOHUADAO INFORMATION TECHNOLOGY Co.,Ltd. |
|
GR01 | Patent grant | ||
GR01 | Patent grant |