CN108563625A - Text analyzing method, apparatus, electronic equipment and computer storage media - Google Patents
Text analyzing method, apparatus, electronic equipment and computer storage media Download PDFInfo
- Publication number
- CN108563625A CN108563625A CN201810007810.8A CN201810007810A CN108563625A CN 108563625 A CN108563625 A CN 108563625A CN 201810007810 A CN201810007810 A CN 201810007810A CN 108563625 A CN108563625 A CN 108563625A
- Authority
- CN
- China
- Prior art keywords
- text
- analyzed
- emotional expression
- attribute
- expression
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
An embodiment of the present invention provides a kind of text analyzing method, apparatus, electronic equipment and computer storage medias.Wherein, text analysis method includes：Obtain text to be analyzed；Wherein, text to be analyzed includes evaluation object；The attribute of evaluation object is identified from text to be analyzed；Emotional expression corresponding with attribute is extracted from text to be analyzed；According to emotional expression, the Sentiment orientation expressed by text to be analyzed is determined.Compared with the prior art is by way of manual analysis text, the embodiment of the present invention is by taking above-mentioned technical proposal, it solves the technical issues of how improving text analyzing efficiency, risk control capability, the ability of market prediction ability and customer relation management can be obviously improved.Especially to the comment data of the fields such as news, video display, product sentence level, analyze more accurate and effective.
Description
Technical field
The present invention relates to technical field of character recognition, more particularly to a kind of text analyzing method, apparatus, electronic equipment and
Computer storage media.
Background technology
Currently, in the fields such as consumer goods industry, service trade, health medical treatment industry, financial services, enterprise, tissue
Equal hope find the opinion of client or the public to its products ＆ services；Meanwhile consumer also is intended to know before it buys commodity
Use the user of the commodity for the evaluation of the commodity.That is, when user is before making decisions, it is desirable to first understand him
The viewpoint (also referred to as emotion) of people, so that understand other people by viewpoint is to product and/or the expressed Sentiment orientation out of service
Positive or negative, and then can be that decision makes anticipation by Sentiment orientation.
However, under the environment of times of information explosion instantly, the prior art is by manually analyzing text, to obtain
Related Sentiment orientation, and then for instructing the decision in actual production, life.
Therefore, there is the defect for causing efficiency low because taking manual type to analyze text in the prior art.
Invention content
The embodiment of the present invention is designed to provide a kind of text analyzing method, apparatus, electronic equipment and computer storage
Medium, to improve the efficiency of text analyzing.
To achieve the goals above, in a first aspect, providing following technical scheme：
A kind of text analyzing method, the method includes：
Obtain text to be analyzed；Wherein, the text to be analyzed includes evaluation object；
The attribute of the evaluation object is identified from the text to be analyzed；
Emotional expression corresponding with the attribute is extracted from the text to be analyzed；
According to the emotional expression, the corresponding Sentiment orientation of the text to be analyzed is determined.
Optionally, described the step of identifying the attribute of the evaluation object from the text to be analyzed, including：
The attribute of the evaluation object is identified from the text to be analyzed using support vector machine classifier.
Optionally, described the step of extracting emotional expression corresponding with the attribute from the text to be analyzed, packet
It includes：
Use condition random field models extract emotional expression corresponding with the attribute from the text to be analyzed.
Optionally, described according to the emotional expression, the step of determining the text to be analyzed corresponding Sentiment orientation, packet
It includes：
Judge that the emotional expression is positive emotion expression or negative emotion expression；
If the emotional expression is positive emotion expression, it is determined that the corresponding Sentiment orientation of the text to be analyzed is just
To Sentiment orientation；
If the emotional expression is negative emotion expression, it is determined that the corresponding Sentiment orientation of the text to be analyzed is negative
To Sentiment orientation.
To achieve the goals above, second aspect additionally provides following technical scheme：
A kind of text analyzing device, described device include：
Acquisition module, for obtaining text to be analyzed；Wherein, the text to be analyzed includes evaluation object；
Identification module, the attribute for identifying the evaluation object from the text to be analyzed；
Extraction module, for extracting emotional expression corresponding with the attribute from the text to be analyzed；
Determining module, for according to the emotional expression, determining the corresponding Sentiment orientation of the text to be analyzed.
Optionally, the identification module is specifically used for：
The attribute of the evaluation object is identified from the text to be analyzed using support vector machine classifier.
Optionally, the extraction module is specifically used for：
Use condition random field models extract emotional expression corresponding with the attribute from the text to be analyzed.
Optionally, the determining module specifically includes：
Judging submodule, for judging that the emotional expression is positive emotion expression or negative emotion expression；
First determination sub-module, if expressed for positive emotion for the emotional expression, it is determined that the text to be analyzed
This corresponding Sentiment orientation is positive Sentiment orientation；
Second determination sub-module, if expressed for negative emotion for the emotional expression, it is determined that the text to be analyzed
This corresponding Sentiment orientation is negative sense Sentiment orientation.
To achieve the goals above, the third aspect additionally provides following technical scheme：
A kind of electronic equipment, including processor, communication interface, memory and communication bus, wherein the processor, institute
It states communication interface and the memory completes mutual communication by the communication bus；
The memory, for storing computer program；
The processor when for executing the program stored on memory, realizes the method and step described in first aspect.
To achieve the goals above, fourth aspect additionally provides following technical scheme：
A kind of computer readable storage medium is stored with computer program in the computer readable storage medium, described
The method and step described in first aspect is realized when computer program is executed by processor.
A kind of text analyzing method, apparatus of offer of the embodiment of the present invention, electronic equipment and computer storage media.Wherein,
Text analysis method includes：Obtain text to be analyzed；Wherein, text to be analyzed includes evaluation object；From text to be analyzed
Identify the attribute of evaluation object；Emotional expression corresponding with attribute is extracted from text to be analyzed；According to emotional expression, really
Sentiment orientation expressed by fixed text to be analyzed.Compared with the prior art is by way of manual analysis text, the present invention is implemented
Example passes through the corresponding emotion table of the attribute for extracting with being identified from text to be analyzed by taking above-mentioned technical proposal
It reaches；Then, judge that the emotional expression gone out expressed by text to be analyzed is inclined to by the emotional expression, to improve text point
The efficiency of analysis.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
It obtains it is clear that being emerged from by implementing the present invention.The purpose of the present invention and other advantages can by specification,
Specifically noted structure is realized and is obtained in claims and attached drawing.
Certainly, implement any of the products of the present invention or method it is not absolutely required at the same reach all the above excellent
Point.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is the flow diagram of the text analyzing method of the embodiment of the present invention；
Fig. 2 be the embodiment of the present invention according to emotional expression, determine the tool of the corresponding Sentiment orientation step of text to be analyzed
Body flow diagram；
Fig. 3 is the structural schematic diagram of the text analyzing device of the embodiment of the present invention；
Fig. 4 is the structural schematic diagram of the determining module of the embodiment of the present invention；
Fig. 5 is the structural schematic diagram of the electronic equipment of the embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Emotion (also referred to as viewpoint) analysis is a subdivision research field of text mining.Specifically, sentiment analysis belongs to letter
The crossing research field of breath retrieval, natural language processing and artificial intelligence.Sentiment analysis is with natural language processing, text analyzing
Extracted automatically with related computer technology or classifying text in emotion.The main study subject of sentiment analysis is internet Zhong Hai
The texts such as the existing comment on commodity of amount, microblogging, blog and forum postings.By analyzing these texts, enterprise can be by a large amount of
Product review obtains the data of consumer feedback in real time and rapidly, to promote product competitiveness, to more fully understand the network user
Behavior provides powerful decision support for enterprise and tissue.From the point of view of the granularity level of analysis, sentiment analysis can be divided into coarseness
With fine granularity step analysis, it is entire text that coarseness step analysis, which has chapter grade and Sentence-level sentiment analysis, hypotheses,
It include only a kind of emotion with sentence.Such hypotheses make the analysis and identification of the two levels not go out in text and sentence not
Same emotional expression object and its corresponding emotion.Therefore, there are the low defects of efficiency for existing coarseness analytical technology.
Requirement with organizations and individuals to sentiment analysis is higher and higher, and the granularity level of sentiment analysis is just gradually to particulate
Spend analysis level development.Fine granularity level sentiment analysis is then the analysis based on evaluation object and its attribute.Based on this, for existing
There are defect existing for technology, the embodiment of the present invention to provide a kind of text analyzing method, as shown in Figure 1, this method includes following step
Rapid S100 to step S130.Wherein：
S100：Obtain text to be analyzed；Wherein, text to be analyzed includes evaluation object.
Wherein, text to be analyzed is the text for expressing viewpoint.Such as：Some or certain objects are seen in the expression such as comment on commodity
The text of method, the text etc. of expression own self emotion (for example, happiness, anger, grief and joy).
Wherein, evaluation object refers to the evaluation object of viewpoint, that is to say a part for the entity, the entity that viewpoint evaluated
Or some or certain attributes of the entity.Wherein, entity can be a product, service, theme, individual, tissue, proposition or
Event etc..In practical applications, e (T, W) can be used to describe an entity.Wherein, T indicates a hierarchical relationship, the level
Relationship includes component, sub-component etc.；Each component or sub-component have the attribute of their own.W indicates an attribute set of e.
For example, by taking " my disagreeable cloudy day " this text as an example, evaluation object is " cloudy day ".For another example, with 6 hand of millet
For machine, which is an entity.6 mobile phone of millet includes attribute set, including：Price, size, weight, screen
The attributes such as curtain, memory, battery.Wherein, battery includes the attributes such as cruising ability, weight, service life.
S110：The attribute of evaluation object is identified from text to be analyzed.
This step can utilize the method based on supervised learning to identify the attribute of evaluation object from text to be analyzed.
In some embodiments, this step specifically includes：It is identified from text to be analyzed using support vector machine classifier
Go out the attribute of evaluation object.
Wherein, sample space is mapped to a higher-dimension by support vector machines by a Nonlinear Mapping (i.e. kernel function),
Even infinite dimensional feature space, by the problem of Nonlinear separability in sample space be converted into higher dimensional space it is linear can
The problem of dividing.Support vector machine classifier can be trained to obtain by training data.In the training process, training corpus is first built
Library；Then, attribute word and the non-attribute word in training corpus are marked manually；Then, word position, semanteme, word is interdependent
Relationship, morphology etc. are used as feature；A sentiment dictionary is built further according to training data；Then, it is based on the sentiment dictionary, is trained
Predefined evaluation object attribute.Finally, accuracy rate, recall rate and F1 (estimating) are calculated by training data, and obtained accordingly
To best support vector machine classifier.
The present embodiment can utilize the word segmentation result and emotion that support vector machine classifier will go out from text dividing to be analyzed
Dictionary is matched, if successful match, it is determined that the word segmentation result is the attribute of evaluation object.
Certainly, there are many noises in text to be analyzed sometimes, or the considerations of for convenient for handling；To evaluation object
Attribute be identified before, text to be analyzed can also be pre-processed.Pretreatment includes formalization processing, at participle
Reason, denoising, best match.
Wherein, formalization processing refers to indicating text with vector space model.The embodiment of the present invention is analysed to text
Originally it is expressed as space vector, in order to carry out sentiment analysis.
For example, for text D=D { t to be analyzed1,w1；t2,w2；......wi, wherein tiIndicate ith feature
, i=1,2 ... n, characteristic item are, for example, word, word, phrase etc.；wiIndicate characteristic item tiWeights, n indicate characteristic item it is total
Number.Wherein, weights can pass through the acquisitions such as word frequency method, information gain method, expectation Cross-Entropy Method.Thus, it is possible to be analysed to text
It is expressed as vector：(w1,w2,...wn)。
Wherein, word segmentation processing refers to by continuous Chinese character sequence, and according to scheduled rule, cutting is word or phrase.In reality
In the application of border, it can consider to carry out using Uni-Gram, two-dimensional grammar model or ternary syntactic model according to actual conditions
Word segmentation processing.
It illustrates below and word segmentation processing is illustrated：
With " I likes playing wechat！" for this text；It is after participle：I, like, play, wechat,！.
Wherein, denoising refers to removing irrelevant information (for example,@lllu111, html) in text to be analyzed, not having point
Class meaning function word (for example, eh, I etc.), to reduce processing complexity.
Wherein, best match is in order to ensure the attribute identified can correctly reflect positive negative sense Sentiment orientation.With " free and unfettered
Outside distant method " for a word, if what is extracted is " carefree " word, positive emotion expression can be erroneously interpreted as, and " carefree method
It is in itself a negative emotion expression outside ".So the matched method of longest may be used in the embodiment of the present invention.
In addition, for commenting on class text, due to it with timeliness n, short text, the spies such as irregularly express, contain much information
Point needs to consider that these factors are pre-processed in practical applications.
S120：Emotional expression corresponding with attribute is extracted from text to be analyzed.
Wherein, emotional expression is embodied by emotion word.For example, for " the screen size reasonable design of iPhone, color
Also gorgeous ", it is therein " reasonable ", " gorgeous " for emotional expression.In order to which the word being analysed in text all remains, this
Full vocabulary method may be used in embodiment, all regard all words as emotional expression.
Wherein, the progress of supervised learning method may be used in extracting method.
In some embodiments, step S120 is specifically included：Use condition random field models are extracted from text to be analyzed
Go out emotional expression corresponding with attribute.
Wherein, conditional random field models are a kind of undirected sequence moulds, pass through different label judgement evaluation objects
Sentiment orientation, i.e., the corresponding Sentiment orientation of text to be analyzed.Wherein, it when conditional random field models carry out text mark, uses
Emotional expression label is：Positive emotional expression starts label (PB), the positive subsequent label of emotional expression (PI), negative sense emotional expression
Start label (NB), the subsequent label of negative sense emotional expression (NI), other (O).
For example, in the case of for given observation sequence x, the condition distribution p (y | x) of hidden sequences y is modeled.
That is being based on training data, observation sequence x is identified by selecting the maximized hidden sequences ies of p (y | x).Therefore, this mistake
Journey can be considered sequence label problem to solve.Table one schematically illustrates emotional expression annotation results.
Table one：
In Table 1, serial number indicates position of each word in sentence；Observation sequence indicates the defeated of conditional random field models
Enter；Annotation results indicate the output of conditional random field models.
Specifically, the present embodiment may comprise steps of：Obtain training corpus and extraction emotional expression.
Wherein, language material is divided into the language material marked and the language material not marked.For example, the language material marked is：To businessman's
Comment, the comment etc. to product, these language materials can determine the Sentiment orientation of client by star or scoring.The language material not marked is such as
The comment etc. of news, these language materials are then needing disaggregated model before or are manually being labeled, and manually to the positive and negative of language material
Tendency.
When extracting emotional expression, by the conditional random field models after training to the attribute of evaluation object in text to be analyzed
Corresponding emotional expression is labeled, then, according to annotation results, you can obtain emotional expression.Wherein, training condition is random
Field model can be based on training corpus, training obtains by building sentiment dictionary.
The extraction performance of conditional random field models can be evaluated by accuracy rate, recall rate and F values (estimating).Its
In：
Wherein, P indicates that accuracy rate, R indicate recall rate；F representing measures.
Certainly, extracted from text to be analyzed in use condition random field models emotional expression corresponding with attribute it
Before, it can also be pre-processed, which can refer to aforementioned preprocessing process, and details are not described herein.
It, can be with as it can be seen that extract emotional expression corresponding with attribute from text to be analyzed by conditional random field models
Excavated from a large amount of text consumer to each attributive character of certain product hold positive attitude still hold reverse side attitude and
To the Sentiment orientation of entire product.
S130：According to emotional expression, the corresponding Sentiment orientation of text to be analyzed is determined.
Specifically, this step determines the corresponding Sentiment orientation of text to be analyzed according to emotion word.
As an example, the embodiment of the present invention can determine Sentiment orientation by way of scoring.With comment on commodity data " product
Matter is pretty good.It is exactly that delivery is too slow, is received with more than ten talentes.Quality is pretty good, has game card to send " for, commentator thinks this
The advantages of product is " quality is pretty good ", the disadvantage is that " delivery is too slow ", overall evaluation is " quality is pretty good, has game card to send ", and
And play the satisfaction of " 5.0 scores " to indicate commentator to the product.The present embodiment can indicate negative sense Sentiment orientation to comment 1~2 point,
3~5 points indicate positive Sentiment orientation.
In some embodiments, as shown in Fig. 2, step S130 specifically comprises the following steps S131 to step S133.
S131：Judge that emotional expression is positive emotion expression or negative emotion expression；
For example, " this chair quality is fine " expresses for positive emotion；" my disagreeable cloudy day " is that negative emotion is expressed.
S132：If emotional expression is positive emotion expression, it is determined that the corresponding Sentiment orientation of text to be analyzed is forward direction
Sentiment orientation.
With "@lllu111:Hello bicycle is really praised very much！Can freely ride 135 days, recommend you use Oh~" for, such as
Fruit emotional expression is " praising ", " recommendations ", it is determined that the corresponding Sentiment orientation of text to be analyzed is forward direction Sentiment orientation.
S133：If emotional expression is negative emotion expression, it is determined that the corresponding Sentiment orientation of text to be analyzed is negative sense
Sentiment orientation.
For example, by taking " I disagreeable cloudy day " as an example, emotional expression is " disagreeable ", it is determined that the corresponding emotion of text to be analyzed is inclined
To for negative sense Sentiment orientation.
Based on the above-mentioned technical proposal, the embodiment of the present invention can be by excavating the text rich in emotional semantic information
The knowledge that magnanimity can be extracted improves text analyzing efficiency, can accurately grasp user's public opinion trend, can be obviously improved risk
The ability of management and control ability, market prediction ability and customer relation management.Especially to the fields such as news, video display, product Sentence-level
Other comment data is analyzed more accurate and effective.
It should be noted that the step in above method embodiment can execute parallel, can also sequentially it execute, the present invention
This is not construed as limiting.And above method embodiment can either hardware be also or the combination of software and hardware gives by software
It realizes.
The embodiment of the present invention also provides a kind of text analyzing device, and text analytical equipment can be used for executing above-mentioned text point
Analyse embodiment of the method.As shown in figure 3, the device includes：
Acquisition module 31, for obtaining text to be analyzed；Wherein, text to be analyzed includes evaluation object；
Identification module 32, the attribute for identifying evaluation object from text to be analyzed；
Extraction module 33, for extracting emotional expression corresponding with attribute from text to be analyzed；
Determining module 34, for according to emotional expression, determining the corresponding Sentiment orientation of text to be analyzed.
The embodiment of the present invention is extracted by taking above-mentioned technical proposal, by extraction module 33 from text to be analyzed：
Emotional expression corresponding with the attribute identified by identification module 32；Then, text to be analyzed is judged by determining module 34
The emotional expression tendency gone out expressed by this, to improve the efficiency of text analyzing.
In some embodiments, above-mentioned identification module 32 is specifically used for：
The attribute of evaluation object is identified from text to be analyzed using support vector machine classifier.
In some embodiments, said extracted module 33 is specifically used for：
Use condition random field models extract emotional expression corresponding with attribute from text to be analyzed.
In some embodiments, as shown in figure 4, determining module 34 specifically includes：
Judging submodule 41, for judging that emotional expression is positive emotion expression or negative emotion expression；
First determination sub-module 42, if expressed for positive emotion for emotional expression, it is determined that text to be analyzed corresponds to
Sentiment orientation be positive Sentiment orientation；
Second determination sub-module 43, if expressed for negative emotion for emotional expression, it is determined that text to be analyzed corresponds to
Sentiment orientation be negative sense Sentiment orientation.
Explanation in relation to above-mentioned text analyzing device can refer to the related description in aforementioned texts analysis method embodiment,
Details are not described herein.
Based on technical concept identical with embodiment of the method, the embodiment of the present invention additionally provides a kind of electronic equipment, such as Fig. 5
It is shown, including processor 51, communication interface 52, memory 53 and communication bus 54, wherein processor 51, communication interface 52 are deposited
Reservoir 53 completes mutual communication by communication bus 54,
Memory 53, for storing computer program；
Processor 51 when for executing the program stored on memory 53, realizes above-mentioned text analyzing embodiment of the method
Described in method and step.
The communication bus 54 that above-mentioned electronic equipment is mentioned can be Peripheral Component Interconnect standard (Peripheral
Component Interconnect, PCI) bus or expanding the industrial standard structure (Extended Industry Standard
Architecture, EISA) bus etc..The communication bus can be divided into address bus, data/address bus, controlling bus etc..For just
It is only indicated with a thick line in expression, Fig. 5, it is not intended that an only bus or a type of bus.
Above-mentioned communication interface 52 is for the communication between above-mentioned electronic equipment and other equipment.
Above-mentioned memory 53 may include random access memory (Random Access Memory, RAM), can also wrap
Include nonvolatile memory (non-volatile memory, NVM), for example, at least a magnetic disk storage.Optionally, it stores
Device can also be at least one storage device for being located remotely from aforementioned processor.
Above-mentioned processor 51 can be general processor, including central processing unit (Central Processing
Unit, abbreviation CPU), network processing unit (Ne twork Processor, NP) etc.；It can also be digital signal processor
(Digital Signal Processing, DSP), application-specific integrated circuit (Application Specific Integrated
Circuit, ASIC), field programmable gate array (Field-Programmable Gate Array, FPGA) or other can
Programmed logic device, discrete gate or transistor logic, discrete hardware components.
Above-mentioned electronic equipment includes but not limited to smart mobile phone, computer, personal digital assistant, wearable device etc..
In the present embodiment, the program stored on memory 53 is executed by processor 51, by from text to be analyzed
Extract emotional expression corresponding with the attribute identified；Then, text institute to be analyzed table is judged by the emotional expression
The emotional expression tendency is reached, to improve the efficiency of text analyzing.
Based on technical concept identical with embodiment of the method, the embodiment of the present invention additionally provides a kind of computer-readable storage
Medium.It is stored with computer program in the computer readable storage medium, is realized when computer program is executed by processor above-mentioned
Method and step described in text analyzing embodiment of the method.
It will be understood by those skilled in the art that above computer program may include some instructions, so that computing device
(for example, personal computer, server etc.) executes the method and step described in any of the above-described text analyzing embodiment of the method.
Above computer readable storage medium storing program for executing can include but is not limited to random access memory (RAM), dynamic random is deposited
Access to memory (DRAM), static RAM (SRAM), read-only memory (ROM), programmable read only memory
(PROM), Erarable Programmable Read only Memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory (example
Such as, NOR type flash memory or NAND-type flash memory), Content Addressable Memory (CAM), polymer memory is (for example, ferroelectric polymers
Memory), phase transition storage, ovonic memory, silicon-oxide-nitride silicon-silica-silicon (Silicon-
Oxide-Nitride-Oxide-Silicon, SONOS) memory, magnetic card or light-card, also or any other appropriate type
Computer readable storage medium.
In the present embodiment, when computer program is executed by processor, by extracting and being identified from text to be analyzed
The corresponding emotional expression of attribute gone out；Then, the emotional expression gone out expressed by text to be analyzed is judged by the emotional expression
Tendency, to improve the efficiency of text analyzing.
It should be noted that herein, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also include other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
Each embodiment in this specification is all made of relevant mode and describes, the highlights of each of the examples are with
The difference of other embodiment, the same or similar parts between the embodiments can be referred to each other.Especially for text point
For analysis apparatus embodiment, since it is substantially similar to the method embodiment, so description is fairly simple, related place is referring to side
The part of method embodiment illustrates.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all
Any modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present invention
It is interior.
Claims (10)
1. a kind of text analyzing method, which is characterized in that the method includes：
Obtain text to be analyzed；Wherein, the text to be analyzed includes evaluation object；
The attribute of the evaluation object is identified from the text to be analyzed；
Emotional expression corresponding with the attribute is extracted from the text to be analyzed；
According to the emotional expression, the corresponding Sentiment orientation of the text to be analyzed is determined.
2. according to the method described in claim 1, it is characterized in that, described identify the evaluation from the text to be analyzed
The step of attribute of object, including：
The attribute of the evaluation object is identified from the text to be analyzed using support vector machine classifier.
3. according to the method described in claim 1, it is characterized in that, described extract and the category from the text to be analyzed
The step of property corresponding emotional expression, including：
Use condition random field models extract emotional expression corresponding with the attribute from the text to be analyzed.
4. according to the method described in claim 1, it is characterized in that, described according to the emotional expression, determine described to be analyzed
The step of text corresponding Sentiment orientation, including：
Judge that the emotional expression is positive emotion expression or negative emotion expression；
If the emotional expression is positive emotion expression, it is determined that the corresponding Sentiment orientation of the text to be analyzed is positive feelings
Sense tendency；
If the emotional expression is negative emotion expression, it is determined that the corresponding Sentiment orientation of the text to be analyzed is negative sense feelings
Sense tendency.
5. a kind of text analyzing device, which is characterized in that described device includes：
Acquisition module, for obtaining text to be analyzed；Wherein, the text to be analyzed includes evaluation object；
Identification module, the attribute for identifying the evaluation object from the text to be analyzed；
Extraction module, for extracting emotional expression corresponding with the attribute from the text to be analyzed；
Determining module, for according to the emotional expression, determining the corresponding Sentiment orientation of the text to be analyzed.
6. device according to claim 5, which is characterized in that the identification module is specifically used for：
The attribute of the evaluation object is identified from the text to be analyzed using support vector machine classifier.
7. device according to claim 5, which is characterized in that the extraction module is specifically used for：
Use condition random field models extract emotional expression corresponding with the attribute from the text to be analyzed.
8. device according to claim 5, which is characterized in that the determining module specifically includes：
Judging submodule, for judging that the emotional expression is positive emotion expression or negative emotion expression；
First determination sub-module, if expressed for positive emotion for the emotional expression, it is determined that the text pair to be analyzed
The Sentiment orientation answered is positive Sentiment orientation；
Second determination sub-module, if expressed for negative emotion for the emotional expression, it is determined that the text pair to be analyzed
The Sentiment orientation answered is negative sense Sentiment orientation.
9. a kind of electronic equipment, which is characterized in that including processor, communication interface, memory and communication bus, wherein described
Processor, the communication interface and the memory complete mutual communication by the communication bus；
The memory, for storing computer program；
The processor when for executing the program stored on memory, realizes any method steps of claim 1-4
Suddenly.
10. a kind of computer readable storage medium, which is characterized in that be stored with computer in the computer readable storage medium
Program realizes claim 1-4 any method and steps when the computer program is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810007810.8A CN108563625A (en) | 2018-01-04 | 2018-01-04 | Text analyzing method, apparatus, electronic equipment and computer storage media |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810007810.8A CN108563625A (en) | 2018-01-04 | 2018-01-04 | Text analyzing method, apparatus, electronic equipment and computer storage media |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108563625A true CN108563625A (en) | 2018-09-21 |
Family
ID=63529715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810007810.8A Pending CN108563625A (en) | 2018-01-04 | 2018-01-04 | Text analyzing method, apparatus, electronic equipment and computer storage media |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108563625A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110175238A (en) * | 2019-05-31 | 2019-08-27 | 杭州网易再顾科技有限公司 | Public sentiment classification method, medium, device and calculating equipment |
| CN110187760A (en) * | 2019-05-14 | 2019-08-30 | 北京百度网讯科技有限公司 | Intelligent interactive method and device |
| CN110309506A (en) * | 2019-05-28 | 2019-10-08 | 北京三快在线科技有限公司 | Statement analytical method, device, electronic equipment and readable storage medium storing program for executing |
| CN112069311A (en) * | 2020-08-04 | 2020-12-11 | 北京声智科技有限公司 | Text extraction method, device, equipment and medium |
| CN112256826A (en) * | 2020-10-19 | 2021-01-22 | 网易（杭州）网络有限公司 | Emotion analysis method, evaluation method and emotion analysis model training method and device |
| CN112308387A (en) * | 2020-10-20 | 2021-02-02 | 深圳思为科技有限公司 | Client intention degree evaluation method and device and cloud server |
| CN113420122A (en) * | 2021-06-24 | 2021-09-21 | 平安科技（深圳）有限公司 | Method, device and equipment for analyzing text and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049435A (en) * | 2013-01-04 | 2013-04-17 | 浙江工商大学 | Text fine granularity sentiment analysis method and text fine granularity sentiment analysis device |
| 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 |
| CN104915443A (en) * | 2015-06-29 | 2015-09-16 | 北京信息科技大学 | Extraction method of Chinese Microblog evaluation object |
| CN106844632A (en) * | 2017-01-20 | 2017-06-13 | 清华大学 | Based on the product review sensibility classification method and device that improve SVMs |
-
2018
- 2018-01-04 CN CN201810007810.8A patent/CN108563625A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049435A (en) * | 2013-01-04 | 2013-04-17 | 浙江工商大学 | Text fine granularity sentiment analysis method and text fine granularity sentiment analysis device |
| 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 |
| CN104915443A (en) * | 2015-06-29 | 2015-09-16 | 北京信息科技大学 | Extraction method of Chinese Microblog evaluation object |
| CN106844632A (en) * | 2017-01-20 | 2017-06-13 | 清华大学 | Based on the product review sensibility classification method and device that improve SVMs |
Non-Patent Citations (1)
| Title |
|---|
| 施寒潇: "细粒度情感分析研究", 《中国博士学位论文全文数据库信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110187760A (en) * | 2019-05-14 | 2019-08-30 | 北京百度网讯科技有限公司 | Intelligent interactive method and device |
| CN110309506A (en) * | 2019-05-28 | 2019-10-08 | 北京三快在线科技有限公司 | Statement analytical method, device, electronic equipment and readable storage medium storing program for executing |
| CN110309506B (en) * | 2019-05-28 | 2021-04-27 | 北京三快在线科技有限公司 | Statement analysis method and device, electronic equipment and readable storage medium |
| CN110175238A (en) * | 2019-05-31 | 2019-08-27 | 杭州网易再顾科技有限公司 | Public sentiment classification method, medium, device and calculating equipment |
| CN112069311A (en) * | 2020-08-04 | 2020-12-11 | 北京声智科技有限公司 | Text extraction method, device, equipment and medium |
| CN112256826A (en) * | 2020-10-19 | 2021-01-22 | 网易（杭州）网络有限公司 | Emotion analysis method, evaluation method and emotion analysis model training method and device |
| CN112308387A (en) * | 2020-10-20 | 2021-02-02 | 深圳思为科技有限公司 | Client intention degree evaluation method and device and cloud server |
| CN112308387B (en) * | 2020-10-20 | 2024-05-14 | 深圳思为科技有限公司 | Customer intention evaluation method and device and cloud server |
| CN113420122A (en) * | 2021-06-24 | 2021-09-21 | 平安科技（深圳）有限公司 | Method, device and equipment for analyzing text and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11620455B2 (en) | Intelligently summarizing and presenting textual responses with machine learning | |
| CN108563625A (en) | Text analyzing method, apparatus, electronic equipment and computer storage media | |
| CN108804512B (en) | Text classification model generation device and method and computer readable storage medium | |
| CN112632385B (en) | Course recommendation method, course recommendation device, computer equipment and medium | |
| CN111260437B (en) | Product recommendation method based on commodity-aspect-level emotion mining and fuzzy decision | |
| CN107992531A (en) | News personalization intelligent recommendation method and system based on deep learning | |
| CN107391760A (en) | User interest recognition methods, device and computer-readable recording medium | |
| CN110825956A (en) | Information flow recommendation method and device, computer equipment and storage medium | |
| CN109933782B (en) | User emotion prediction method and device | |
| CN112347367A (en) | Information service providing method, information service providing device, electronic equipment and storage medium | |
| Aralikatte et al. | Fault in your stars: an analysis of android app reviews | |
| Alamsyah et al. | Measuring Public Sentiment Towards Services Level in Online Forum Using Naive Bayes Classifier and Word Cloud | |
| CN117216275A (en) | Text processing method, device, equipment and storage medium | |
| CN111950265A (en) | Domain lexicon construction method and device | |
| CN110647504B (en) | Method and device for searching judicial documents | |
| CN112487132A (en) | Keyword determination method and related equipment | |
| CN114780724A (en) | Case classification method and device, computer equipment and storage medium | |
| CN110717008B (en) | Search result ordering method and related device based on semantic recognition | |
| CN113837836A (en) | Model recommendation method, device, equipment and storage medium | |
| CN113392218A (en) | Training method of text quality evaluation model and method for determining text quality | |
| CN109241993B (en) | Evaluation object emotion classification method and device integrating user and overall evaluation information | |
| Arora et al. | Evaluation Of Product Reviews Using Deep Learning Classifier Models | |
| CN112732908A (en) | Test question novelty evaluation method and device, electronic equipment and storage medium | |
| CN112287215A (en) | Intelligent employment recommendation method and device | |
| CN113536802A (en) | Method, device, equipment and storage medium for judging emotion of text data in languages |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180921 |