CN104573030A - Textual emotion prediction method and device - Google Patents
Textual emotion prediction method and device Download PDFInfo
- Publication number
- CN104573030A CN104573030A CN201510018521.4A CN201510018521A CN104573030A CN 104573030 A CN104573030 A CN 104573030A CN 201510018521 A CN201510018521 A CN 201510018521A CN 104573030 A CN104573030 A CN 104573030A
- Authority
- CN
- China
- Prior art keywords
- text
- mood
- sorted
- emotional
- word
- 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
Landscapes
- Machine Translation (AREA)
Abstract
The application provides a textual emotion prediction method and a textual emotion prediction device. The textual emotion prediction method includes: establishing an emotion matching knowledge base, matching texts to be classified with the emotion matching knowledge base so as to obtain emotional characteristics of the texts to be classified, and classifying the emotional characteristics of the texts to be classified so as to obtain an emotion classification result of the texts. According to the textual emotion prediction method and the textual emotion prediction device, the texts to be classified are automatically studied and classified on the basis of manually establishing the emotion matching knowledge base with action and object markings, accuracy and efficiency of emotional prediction of a reader are improved, and large scale text corpora processing demands are met.
Description
Technical field
The application relates to a kind of text emotional prediction method and device.
Background technology
The important applied field that robotization mood analysis and prediction is the artificial intelligence technology such as natural language processing, mood calculating is carried out to text data, its fundamental purpose is application natural language processing technique and mood computing technique, predicts the mood classification that may trigger after readers ' reading text.Along with the development of infotech and deepening constantly of information system application scope, carrying out robotization mood analysis and prediction to text data is the important technique measure improving enterprises production efficiency and competitive edge.
For needing to process the industry of a large amount of text data, along with the developing of all kinds of digitized business, in vast as the open sea data, how to predict the mood that may trigger of user to product, the interest of excavation user and demand are all the direct challenges faced by large data processing industries.But, in prior art, also do not have a kind ofly effectively to predict the method for reader to the reading mood of text, if carry out classification prediction with the mood classification comprised of the mode of artificial prediction processing to text, for extensive text, great manpower can be wasted, cause classification effectiveness low.
Summary of the invention
The application provides a kind of text emotional prediction method and device, can improve the accuracy and efficiency of reader's emotional prediction, meets the demand of extensive corpus of text process.
According to the first aspect of the application, the application provides a kind of text emotional prediction method, comprise: build and be used for mating knowledge base with the mood that text to be sorted carries out mating, described mood coupling knowledge base comprises the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action, and described degrees of emotion is for representing the degree of strength of described mood classification; Described text to be sorted is mated knowledge base with described mood mate, obtain the emotional characteristics of text to be sorted, described emotional characteristics comprises the mood classification of described text to be sorted, degrees of emotion, agent and word denoting the receiver of an action; The emotional characteristics of described text to be sorted is classified, obtains the mood classification results of text.
According to the second aspect of the application, the application provides a kind of text emotional prediction device, comprise: construction unit, for building for mating knowledge base with the mood that text to be sorted carries out mating, described mood coupling knowledge base comprises the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action, and described degrees of emotion is for representing the degree of strength of described mood classification; Matching unit, mates for described text to be sorted is mated knowledge base with described mood, obtains the emotional characteristics of text to be sorted, and described emotional characteristics comprises the mood classification of described text to be sorted, degrees of emotion, agent and word denoting the receiver of an action; Taxon, for being classified by the emotional characteristics of described text to be sorted, obtains the mood classification results of text.
The text emotional prediction method that the application provides and device, build mood coupling knowledge base, text to be sorted is mated knowledge base with mood and mates, obtain the emotional characteristics of text to be sorted, the emotional characteristics of described text to be sorted is classified, obtains the mood classification results of text.There is action mate on the basis of knowledge base artificial constructed with the mood of object marking, text automatic learning to be sorted is classified, improve the accuracy and efficiency of reader's emotional prediction, meet the demand of extensive corpus of text process.
Accompanying drawing explanation
Fig. 1 is the method flow diagram of the embodiment of the present invention;
Fig. 2 is the another kind of method flow diagram of the embodiment of the present invention;
Fig. 3 is the apparatus structure schematic diagram of inventive embodiments;
Fig. 4 is the another kind of apparatus structure schematic diagram of inventive embodiments.
Embodiment
By reference to the accompanying drawings the present invention is described in further detail below by embodiment.
In the embodiment of the present application, a kind of text emotional prediction method and device are provided, the accuracy and efficiency of reader's emotional prediction can be improved, meet the demand of extensive corpus of text process.
Embodiment one:
Please refer to Fig. 1, Fig. 1 is the method flow diagram of the embodiment of the present invention one.As shown in Figure 1, a kind of text emotional prediction method, can comprise the following steps:
101, mood coupling knowledge base is built.
Mood coupling knowledge base comprises the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action, and mood coupling knowledge base is used for mating with text to be sorted.The degrees of emotion of the present embodiment is for representing the degree of strength of described mood classification.
Above-mentioned sentence can be chosen from conventional Chinese dictionary, Chinese language knowledge storehouse, the large-scale corpus such as microblogging, news.
The sentence of the embodiment of the present application is for mating with text to be sorted, and the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action can be implemented by the mode of artificial mark.
102, text to be sorted is mated knowledge base with mood to mate, obtain the emotional characteristics of text to be sorted.
Emotional characteristics comprises the mood classification of text to be sorted, degrees of emotion, agent and word denoting the receiver of an action.
The mood classification of text to be sorted and degrees of emotion obtain after mating knowledge base with mood and mating.Concrete matching process can be mated by analysis tools such as parsing tree and dependency analysis trees.
103, the emotional characteristics of text to be sorted is classified.
The mood classification results of text can be obtained after classification.The result of classifying in the present embodiment shows as the probability of happening of mood classification that text packets to be sorted contains and each described mood classification.In an embodiment, the probability of happening of mood classification can also be sorted, finally the result after sequence can also be shown to user and use, thus reach the object that the mood treating classifying text carries out classifying and predicting.
The text emotional prediction method that the application provides, build mood coupling knowledge base, text to be sorted is mated knowledge base with mood and mates, obtain the emotional characteristics of text to be sorted, the emotional characteristics of described text to be sorted is classified, obtains the mood classification results of text.There is action mate on the basis of knowledge base artificial constructed with the mood of object marking, text automatic learning to be sorted is classified, improve the accuracy and efficiency of reader's emotional prediction, meet the demand of extensive corpus of text process.
Embodiment two:
Please refer to Fig. 2, Fig. 2 is the method flow diagram of the embodiment of the present invention two.As shown in Figure 2, a kind of text emotional prediction method, can comprise the following steps:
201A, from commonsense knowledge base, choose verb and/or adjective, be labeled as mood trigger word to be matched.
In the research of artificial intelligence, common sense knowledge refers to the information of the fact and the ordinary people's precognition collected.Commonsense knowledge base is a kind of knowledge base storing common sense knowledge, common commonsense knowledge base comprises WordNet (word net, a kind of commonsense knowledge base of dictionary formula), CYC (Cycorp company develops and the database safeguarded), Thought Treasure (thought collect, one relates to natural language processing commonsense knowledge base), SemanticWeb (semantic net will have the network of domain knowledge base and commonsense knowledge base while a kind of future), Open MindCommon Sense (opening heart commonsense knowledge base) etc.These all belong to general commonsense knowledge base, cannot directly apply in reader's emotional prediction.A lot of emotion expression services of the mankind be implicit, there is no mood word, but these emotion expression services can carry out analysis based on the common sense knowledge hidden in brain obtains.
In the embodiment of the present application, from commonsense knowledge base, choose multiple verb or adjective, or, choose verb and adjective simultaneously.Concrete mode of choosing is arranged according to actual needs.If the verb in English general knowledge storehouse and adjective, then mechanical translation is adopted to correct the method combined translate into Chinese verb and adjective with artificial.When choosing, note the uncommon word that removal dittograph and part are of little use and the word of part archaic Chinese.
Above-described verb and adjective meaning can not certain moods of direct representation, and usually, verb and adjective can trigger the appearance of certain mood.As: " beating " is a verb, and the mood classification that may trigger reader is: indignation, abhor, frightened, censure.These adjectives and verb mark out by the present embodiment step, as mood trigger word to be matched.
201B, from corpus, select the sentence including mood trigger word to be matched.
The sentence with above-mentioned mood trigger word to be matched can be searched from the corpus such as extensive microblogging, news material, blog.
201C, according to mood trigger word to be matched, the mark mood classification of sentence and the agent of degrees of emotion and mood trigger word to be matched and word denoting the receiver of an action.
Degrees of emotion is for representing the degree of strength of degrees of emotion.
In the present embodiment, degrees of emotion 1,3,5,7,9 five grade marks, and the intensity of this mood classification of the larger expression of numeral is stronger.Same word can belong to multiple mood classification.Mood classification is divided into 7 large classifications and 21 little classifications, as shown in table 1
| Numbering | Emotion class (large class) | Emotion class (group) |
| 1 | Happy | Happy, feel at ease |
| 2 | Good | Respect, praise, believe, like, wish |
| 3 | Anger | Indignation |
| 4 | Sorrow | Sad, disappointed, remorse, think of |
| 5 | Fear | Unbearably, frightened, shy |
| 6 | Dislike | Unhappy, abhor, censure, envy, suspect |
| 7 | Frightened | In surprise |
Table 1
Suppose in corpus, the sentence that have chosen " police stops thief " is example, " grab " as mood trigger word to be matched, mark " police stops thief " mood classification in praise of, feel at ease, and the degrees of emotion that mark is praised is 3, the degrees of emotion of feeling at ease is 5, and agent is police, and word denoting the receiver of an action is thief.
Above step 201A-201C is a kind of specific implementation process building mood coupling knowledge base, and this mood coupling knowledge base is used for mating with text to be sorted.After building mood coupling knowledge base, namely described text to be sorted can be mated knowledge base with described mood and mate, concrete implementation process can as step 202A-202D.
202A, mark out mood trigger word in text to be sorted by the analysis tool preset, and the agent of mood trigger word and word denoting the receiver of an action.
The analysis tool preset comprises: parsing tree and dependency analysis tree.By analysis tool, the verb and/or adjective that reader's mood in text to be sorted, can be triggered can be gone out by automatic marking, be called mood trigger word.Meanwhile, agent corresponding to mood trigger word and word denoting the receiver of an action can also be marked out.
202B, in mood coupling knowledge base, search the sentence of coupling.
The mood trigger word to be matched that the sentence of coupling carries is identical with the mood trigger word of text to be sorted.Also namely, in mood coupling knowledge base, search the sentence with the mood trigger word in text to be sorted, be the sentence of coupling.
202C, from coupling sentence filter out training example sentence.
The agent of training example sentence is identical with the meaning of word denoting the receiver of an action with the agent of text to be sorted or close with word denoting the receiver of an action.
Be in a bad mood in the sentence of trigger word at band, filter out the sentence that agent is identical or close with word denoting the receiver of an action with the agent in text to be sorted with word denoting the receiver of an action, namely can be used as training example sentence.
202D, using the training mood classification of example sentence and emotional intensity as the mood classification of text to be sorted and emotional intensity, in conjunction with agent and the word denoting the receiver of an action of text to be sorted, obtain the emotional characteristics of text to be sorted.
After for text matches to be sorted to suitable training example sentence, using the mood classification in training example sentence and emotional intensity as the mood classification of text to be sorted and emotional intensity.The agent of text to be sorted and word denoting the receiver of an action are combined with mood classification and emotional intensity, as the emotional characteristics of text to be sorted by the present embodiment.
Suppose that text to be sorted is that " pedlar beats in municipal administration." specifically implementation step is as follows: the mood trigger word of text to be sorted is: " beating ", and agent is " municipal administration ", and word denoting the receiver of an action is " pedlar ".Mate in knowledge base to search with the mood of object marking and " beat " identical entry with verb having action, find agent more same or similar with " municipal administration ", word denoting the receiver of an action and " pedlar " same or analogous entry, the reader's mood class (group) finding this entry to trigger is " indignation ", " abhoing ", " censuring " and " fear ", and corresponding emotional intensity is respectively 9/7/7/5.Therefore, the feature of text to be sorted is as shown in table 2:
| Mood classification (group) | Degrees of emotion | Mood trigger word | Agent | Word denoting the receiver of an action | Mood word |
| Indignation/abhor/censure/frightened | 9/7/7/5 | Beat | Municipal administration | Pedlar | Nothing |
Table 2
203, the emotional characteristics treating classifying text by machine learning classification method is classified.
The probability of happening of mood classification that text packets to be sorted contains and each mood classification is obtained after classification.Machine learning classification method comprises: decision tree, naive Bayesian, support vector machine.
Using the feature that the emotional characteristics in table 2 learns as machine learning classification method, adopt machine learning classification method to treat classifying text as decision tree, naive Bayesian, support vector machine etc. and classify, obtain classification results as shown in table 3:
Table 3
In an embodiment of the application, can also step be comprised:
204, the probability of happening of each mood classification is exported, be shown to user.
The above results sequence output obtains the classification that text to be sorted " pedlar beats in municipal administration " may trigger reader's mood to be had: indignation, abhor, to censure and frightened.
Embodiment three:
Accordingly, the application also provides a kind of text emotional prediction device, please refer to Fig. 3, and Fig. 3 is apparatus structure schematic diagram of the invention process.As shown in Figure 3, text emotional prediction device can comprise:
Construction unit 30, for building for mating knowledge base with the mood that text to be sorted carries out mating, described mood coupling knowledge base comprises the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action, and described degrees of emotion is for representing the degree of strength of described mood classification;
Matching unit 31, mates for described text to be sorted is mated knowledge base with described mood, obtains the emotional characteristics of text to be sorted, and described emotional characteristics comprises the mood classification of described text to be sorted, degrees of emotion, agent and word denoting the receiver of an action;
Taxon 32, for being classified by the emotional characteristics of described text to be sorted, obtains the mood classification results of text.
Please also refer to Fig. 4, in an embodiment, the construction unit 30 of the text emotional prediction device of the application comprises:
Mark unit 300, for choosing verb and/or adjective from commonsense knowledge base, is labeled as mood trigger word to be matched.
Choose unit 301, for selecting the sentence including described mood trigger word to be matched from corpus.
Dispensing unit 302, for according to described mood trigger word to be matched, configure mood classification and the degrees of emotion of described sentence, described degrees of emotion is for representing the degree of strength of described degrees of emotion, and, second mark unit, for marking agent and the word denoting the receiver of an action of described mood trigger word to be matched.
In an embodiment, the matching unit 31 of the text emotional prediction device of the application comprises:
Analyze mark unit 310, for being marked out the mood trigger word in described text to be sorted by default analysis tool, and the agent of described mood trigger word and word denoting the receiver of an action.
Search unit 311, for searching the sentence of coupling in described mood coupling knowledge base, the mood trigger word described to be matched that the sentence of described coupling carries is identical with the described mood trigger word of described text to be sorted.
Screening unit 312, for filtering out training example sentence in the sentence from described coupling, the agent of described training example sentence is identical with the meaning of word denoting the receiver of an action or close with the agent of described text to be sorted with word denoting the receiver of an action.
Allocation units 313, for using the mood classification of described training example sentence and emotional intensity as the mood classification of described text to be sorted and emotional intensity, in conjunction with agent and the word denoting the receiver of an action of described text to be sorted, obtain the emotional characteristics of described text to be sorted.
In an embodiment, taxon 32 specifically for: classified by the emotional characteristics of machine learning classification method to described text to be sorted, obtain the probability of happening of mood classification that described text packets to be sorted contains and each described mood classification, as the mood classification results of text.
The text emotional prediction device that the application provides, for building mood coupling knowledge base, text to be sorted being mated knowledge base with mood and mates, obtaining the emotional characteristics of text to be sorted, the emotional characteristics of described text to be sorted is classified, obtains the mood classification results of text.There is action mate on the basis of knowledge base artificial constructed with the mood of object marking, text automatic learning to be sorted is classified, improve the accuracy and efficiency of reader's emotional prediction, meet the demand of extensive corpus of text process.
Above content is in conjunction with concrete embodiment further description made for the present invention, can not assert that specific embodiment of the invention is confined to these explanations.For general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, some simple deduction or replace can also be made.
Claims (10)
1. a text emotional prediction method, is characterized in that, comprising:
Build and be used for mate knowledge base with the mood that text to be sorted carries out mating, described mood is mated knowledge base and is comprised the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action, and described degrees of emotion is for representing the degree of strength of described mood classification;
Described text to be sorted is mated knowledge base with described mood mate, obtain the emotional characteristics of text to be sorted, described emotional characteristics comprises the mood classification of described text to be sorted, degrees of emotion, agent and word denoting the receiver of an action;
The emotional characteristics of described text to be sorted is classified, obtains the mood classification results of text.
2. text emotional prediction method as claimed in claim 1, is characterized in that, described structure is used for mating knowledge base with the mood that text to be sorted carries out mating and comprises:
From commonsense knowledge base, choose verb and/or adjective, be labeled as mood trigger word to be matched;
The sentence including described mood trigger word to be matched is selected from corpus;
According to described mood trigger word to be matched, mark mood classification and the degrees of emotion of described sentence, and, the agent of described mood trigger word to be matched and word denoting the receiver of an action.
3. text emotional prediction method as claimed in claim 1, is characterized in that, describedly described text to be sorted is mated knowledge base with described mood mates, and the emotional characteristics obtaining text to be sorted comprises:
Mood trigger word in described text to be sorted is marked out by the analysis tool preset, and the agent of described mood trigger word and word denoting the receiver of an action;
In described mood coupling knowledge base, search the sentence of coupling, the mood trigger word described to be matched that the sentence of described coupling carries is identical with the described mood trigger word of described text to be sorted;
From the sentence of described coupling, filter out training example sentence, the agent of described training example sentence is identical with the meaning of word denoting the receiver of an action or close with the agent of described text to be sorted with word denoting the receiver of an action;
Using the mood classification of described training example sentence and emotional intensity as the mood classification of described text to be sorted and emotional intensity, in conjunction with agent and the word denoting the receiver of an action of described text to be sorted, obtain the emotional characteristics of described text to be sorted.
4. text emotional prediction method as claimed in claim 3, it is characterized in that, described default analysis tool comprises:
Parsing tree and dependency analysis tree.
5. text emotional prediction method as claimed in claim 1, it is characterized in that, the described emotional characteristics by described text to be sorted is classified, and the mood classification results obtaining text comprises:
Classified by the emotional characteristics of machine learning classification method to described text to be sorted, obtain the probability of happening of mood classification that described text packets to be sorted contains and each described mood classification, as the mood classification results of text.
6. a text emotional prediction device, is characterized in that, comprising:
Construction unit, for building for mating knowledge base with the mood that text to be sorted carries out mating, described mood coupling knowledge base comprises the mood classification of sentence, degrees of emotion, agent and word denoting the receiver of an action, and described degrees of emotion is for representing the degree of strength of described mood classification;
Matching unit, mates for described text to be sorted is mated knowledge base with described mood, obtains the emotional characteristics of text to be sorted, and described emotional characteristics comprises the mood classification of described text to be sorted, degrees of emotion, agent and word denoting the receiver of an action;
Taxon, for being classified by the emotional characteristics of described text to be sorted, obtains the mood classification results of text.
7. text emotional prediction device as claimed in claim 6, it is characterized in that, described construction unit comprises:
Mark unit, for choosing verb and/or adjective from commonsense knowledge base, is labeled as mood trigger word to be matched;
Choose unit, for selecting the sentence including described mood trigger word to be matched from corpus;
Dispensing unit, for according to described mood trigger word to be matched, configure mood classification and the degrees of emotion of described sentence, described degrees of emotion is for representing the degree of strength of described degrees of emotion, and, second mark unit, for marking agent and the word denoting the receiver of an action of described mood trigger word to be matched.
8. text emotional prediction device as claimed in claim 6, it is characterized in that, described matching unit comprises:
Analyze mark unit, for being marked out the mood trigger word in described text to be sorted by default analysis tool, and the agent of described mood trigger word and word denoting the receiver of an action;
Search unit, for searching the sentence of coupling in described mood coupling knowledge base, the mood trigger word described to be matched that the sentence of described coupling carries is identical with the described mood trigger word of described text to be sorted;
Screening unit, for filtering out training example sentence in the sentence from described coupling, the agent of described training example sentence is identical with the meaning of word denoting the receiver of an action or close with the agent of described text to be sorted with word denoting the receiver of an action;
Allocation units, for using the mood classification of described training example sentence and emotional intensity as the mood classification of described text to be sorted and emotional intensity, in conjunction with agent and the word denoting the receiver of an action of described text to be sorted, obtain the emotional characteristics of described text to be sorted.
9. text emotional prediction device as claimed in claim 8, it is characterized in that, described default analysis tool comprises:
Parsing tree and dependency analysis tree.
10. text emotional prediction device as claimed in claim 6, it is characterized in that, described taxon specifically for: classified by the emotional characteristics of machine learning classification method to described text to be sorted, obtain the probability of happening of mood classification that described text packets to be sorted contains and each described mood classification, as the mood classification results of text.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510018521.4A CN104573030B (en) | 2015-01-14 | 2015-01-14 | A kind of text emotional prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510018521.4A CN104573030B (en) | 2015-01-14 | 2015-01-14 | A kind of text emotional prediction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104573030A true CN104573030A (en) | 2015-04-29 |
| CN104573030B CN104573030B (en) | 2017-12-12 |
Family
ID=53089092
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510018521.4A Active CN104573030B (en) | 2015-01-14 | 2015-01-14 | A kind of text emotional prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104573030B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106055596A (en) * | 2016-05-23 | 2016-10-26 | 南开大学 | Multi-tag on-line news reader emotion prediction method |
| CN106096717A (en) * | 2016-06-03 | 2016-11-09 | 北京光年无限科技有限公司 | Information processing method and system towards intelligent robot |
| CN107220591A (en) * | 2017-04-28 | 2017-09-29 | 哈尔滨工业大学深圳研究生院 | Multi-modal intelligent mood sensing system |
| CN108520030A (en) * | 2018-03-27 | 2018-09-11 | 深圳中兴网信科技有限公司 | File classification method, Text Classification System and computer installation |
| CN109730699A (en) * | 2018-12-28 | 2019-05-10 | 苏州泓邃生物科技有限公司 | A kind of emotional prediction method based on vital sign data |
| CN111597301A (en) * | 2020-04-24 | 2020-08-28 | 北京百度网讯科技有限公司 | Text prediction method and device and electronic equipment |
| US11272888B1 (en) | 2021-06-22 | 2022-03-15 | Nft Ip Holdings Llc | Devices and systems that measure, quantify, compare, package, and capture human content in databases |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102200969A (en) * | 2010-03-25 | 2011-09-28 | 日电（中国）有限公司 | Text sentiment polarity classification system and method based on sentence sequence |
| CN102930042A (en) * | 2012-11-13 | 2013-02-13 | 五邑大学 | Tendency text automatic classification system and achieving method of the same |
| CN103646088A (en) * | 2013-12-13 | 2014-03-19 | 合肥工业大学 | Product comment fine-grained emotional element extraction method based on CRFs and SVM |
| CN104239554A (en) * | 2014-09-24 | 2014-12-24 | 南开大学 | Cross-domain and cross-category news commentary emotion prediction method |
-
2015
- 2015-01-14 CN CN201510018521.4A patent/CN104573030B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102200969A (en) * | 2010-03-25 | 2011-09-28 | 日电（中国）有限公司 | Text sentiment polarity classification system and method based on sentence sequence |
| CN102930042A (en) * | 2012-11-13 | 2013-02-13 | 五邑大学 | Tendency text automatic classification system and achieving method of the same |
| CN103646088A (en) * | 2013-12-13 | 2014-03-19 | 合肥工业大学 | Product comment fine-grained emotional element extraction method based on CRFs and SVM |
| CN104239554A (en) * | 2014-09-24 | 2014-12-24 | 南开大学 | Cross-domain and cross-category news commentary emotion prediction method |
Non-Patent Citations (3)
| Title |
|---|
| 何跃 等: "《中文微博的情绪识别与分类研究》", 《情报杂志》 * |
| 张莹: ""在线新闻评论的情感分析研究"", 《万方》 * |
| 陈涛 等: ""一种基于情感句模的文本情感分类方法"", 《中文信息学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106055596A (en) * | 2016-05-23 | 2016-10-26 | 南开大学 | Multi-tag on-line news reader emotion prediction method |
| CN106096717A (en) * | 2016-06-03 | 2016-11-09 | 北京光年无限科技有限公司 | Information processing method and system towards intelligent robot |
| CN106096717B (en) * | 2016-06-03 | 2018-08-14 | 北京光年无限科技有限公司 | Information processing method towards intelligent robot and system |
| CN107220591A (en) * | 2017-04-28 | 2017-09-29 | 哈尔滨工业大学深圳研究生院 | Multi-modal intelligent mood sensing system |
| CN108520030A (en) * | 2018-03-27 | 2018-09-11 | 深圳中兴网信科技有限公司 | File classification method, Text Classification System and computer installation |
| CN108520030B (en) * | 2018-03-27 | 2022-02-11 | 深圳中兴网信科技有限公司 | Text classification method, text classification system and computer device |
| CN109730699A (en) * | 2018-12-28 | 2019-05-10 | 苏州泓邃生物科技有限公司 | A kind of emotional prediction method based on vital sign data |
| CN111597301A (en) * | 2020-04-24 | 2020-08-28 | 北京百度网讯科技有限公司 | Text prediction method and device and electronic equipment |
| US11272888B1 (en) | 2021-06-22 | 2022-03-15 | Nft Ip Holdings Llc | Devices and systems that measure, quantify, compare, package, and capture human content in databases |
| US11382570B1 (en) | 2021-06-22 | 2022-07-12 | Nft Ip Holdings Llc | Encrypted securitized databases for devices and systems that measure, quantify, compare, package, and capture data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104573030B (en) | 2017-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107861939B (en) | Domain entity disambiguation method fusing word vector and topic model | |
| CN104573030A (en) | Textual emotion prediction method and device | |
| CN109635297B (en) | Entity disambiguation method and device, computer device and computer storage medium | |
| Abujar et al. | A heuristic approach of text summarization for Bengali documentation | |
| CN111046656B (en) | Text processing method, text processing device, electronic equipment and readable storage medium | |
| Jha et al. | Homs: Hindi opinion mining system | |
| CN108549723B (en) | Text concept classification method and device and server | |
| He et al. | Language feature mining for music emotion classification via supervised learning from lyrics | |
| Ogada et al. | N-gram based text categorization method for improved data mining | |
| Chang et al. | A METHOD OF FINE-GRAINED SHORT TEXT SENTIMENT ANALYSIS BASED ON MACHINE LEARNING. | |
| CN108960772A (en) | Enterprise's evaluation householder method and system based on deep learning | |
| Samih et al. | Enhanced sentiment analysis based on improved word embeddings and XGboost. | |
| Philemon et al. | A machine learning approach to multi-scale sentiment analysis of amharic online posts | |
| CN106598952A (en) | System for detecting Chinese fuzzy constraint information scope based on convolutional neural network | |
| Sharma et al. | Lexicon a linguistic approach for sentiment classification | |
| CN109241521B (en) | Scientific literature high-attention sentence extraction method based on citation relation | |
| Dawar et al. | Text categorization by content using Naïve Bayes approach | |
| Munarko et al. | Named entity recognition model for Indonesian tweet using CRF classifier | |
| Lamparter et al. | Knowledge extraction from classification schemas | |
| CN110705285A (en) | Government affair text subject word bank construction method, device, server and readable storage medium | |
| Li et al. | Feature-specific named entity recognition in software development social content | |
| CN109858035A (en) | A kind of sensibility classification method, device, electronic equipment and readable storage medium storing program for executing | |
| Agirrezabal | KU-CST at the Profiling Fake News spreaders Shared Task. | |
| CN110069780B (en) | Specific field text-based emotion word recognition method | |
| Katsurai | Using word embeddings for library and information science research: A short survey |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |