CN108228563A - A kind of user comment analysis method and device - Google Patents
A kind of user comment analysis method and device Download PDFInfo
- Publication number
- CN108228563A CN108228563A CN201711484686.6A CN201711484686A CN108228563A CN 108228563 A CN108228563 A CN 108228563A CN 201711484686 A CN201711484686 A CN 201711484686A CN 108228563 A CN108228563 A CN 108228563A
- Authority
- CN
- China
- Prior art keywords
- user comment
- word
- user
- classification
- comment
- 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
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
Abstract
The invention discloses a kind of user comment analysis method and device, including：Word segmentation processing is carried out to user comment to be analyzed, every user comment is resolved into word；Using preset time as the period, the word frequency of each word in nearest two periods is counted respectively；Export the word that word frequency difference absolute value in two periods is more than first threshold.Solve lack in the prior art it is a kind of user comment is effectively analyzed method the technical issues of.
Description
Technical field
The present invention relates to technical field of data processing more particularly to a kind of user comment analysis method and devices.
Background technology
With internet and the development of e-commerce, more and more clients select shopping online.After net purchase, client
Often Experience Degree etc. in product and purchase product process is evaluated, so comment is analyzed can not only by businessman
Businessman is helped to know public praise of itself product in market, also can understand the deficiency of itself product in time, and then take corrective measure,
It can be the demand that businessman understands customer in time simultaneously, decision support is provided to new product development etc..
Lack a kind of method effectively analyzed user comment in the prior art.
Invention content
The present invention provides a kind of user comment analysis method and device, solve lack in the prior art it is a kind of to user
The technical issues of commenting on the method effectively analyzed.
The present invention provides a kind of user comment analysis method, including：
Word segmentation processing is carried out to user comment to be analyzed, every user comment is resolved into word；
Using preset time as the period, the word frequency of each word in nearest two periods is counted respectively；
Export the word that word frequency difference absolute value in two periods is more than first threshold.
Preferably,
The user comment analysis method further includes：
Classification processing is carried out to the user comment to be analyzed, the user comment is classified as corresponding classification, every
The user comment corresponds to one or more classifications.
Preferably,
The classification processing that user comment every described carries out is specifically included：
According to the preset classification keyword included in the user comment and the corresponding weight of each classification keyword,
Calculate the user comment weight and, by the user comment be included into second threshold be less than the weight sum classification, wherein
Each classification corresponds to a second threshold, and corresponding one or more classification keywords.
Preferably,
Before word segmentation processing and classification processing is carried out to the user comment to be analyzed, further include：
Select from all user comments crawled negative user comment according to emotion, and by the negative user
Comment is as the user comment to be analyzed.
Preferably,
The user comment analysis method further includes：
According to the preset emotion word included in the user comment, the preset interference word included, it is each described in
The preset corresponding weight of emotion word and the length of the user comment carry out emotion color to user comment to be analyzed
Coloured silk divides.
The present invention provides a kind of user comment analytical equipment, including：
For carrying out word segmentation processing to user comment to be analyzed, every user comment is resolved into for word-dividing mode
Word；
Word frequency statistics module, for using preset time as the period, counting the word of each word in nearest two periods respectively
Frequently；
Output module, for exporting the word that word frequency difference absolute value in two periods is more than first threshold.
Preferably,
The user comment analytical equipment, further includes：
The user comment for carrying out classification processing to the user comment to be analyzed, is classified as phase by sort module
The classification answered, every user comment correspond to one or more classifications.
Preferably,
The sort module specifically includes the classification processing that user comment every described carries out：
According to the preset classification keyword included in the user comment and the corresponding weight of each classification keyword,
Calculate the user comment weight and, by the user comment be included into second threshold be less than the weight sum classification, wherein
Each classification corresponds to a second threshold, and corresponding one or more classification keywords.
Preferably,
The user comment analytical equipment, further includes：
Module is chosen in comment, is commented for selecting negative user from all user comments crawled according to emotion
By, and the negative user is commented on as the user comment to be analyzed.
Preferably,
The user comment analytical equipment, further includes：
Emotion division module, for according to the preset emotion word that is included in the user comment, include
The length of preset interference word, each preset corresponding weight of emotion word and the user comment is to be analyzed
User comment carry out emotion division.
As can be seen from the above technical solutions, the present invention has the following advantages：
Word segmentation processing first is carried out to user comment to be analyzed, every user comment is then resolved into word, then
Using preset time as the period, the word frequency of each word in nearest two periods is counted respectively, so as to export word frequency in two periods
The word that poor absolute value is more than first threshold is checked for analysis, because user comment is included much to businessman without reference to value
Word, such as " product " this word, inventor it has been investigated that this word without reference to value within each period
Word frequency be roughly the same, so the present invention can filter out the word for having reference value to businessman, solve the prior art
In lack it is a kind of user comment is effectively analyzed method the technical issues of.
Description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, 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 having to pay creative labor, may be used also
To obtain other attached drawings according to these attached drawings.
Fig. 1 is a kind of flow diagram of the first embodiment of user comment analysis method provided by the invention；
Fig. 2 is a kind of flow diagram of the second embodiment of user comment analysis method provided by the invention；
Fig. 3 is a kind of structure diagram of the first embodiment of user comment analytical equipment provided by the invention；
Fig. 4 is a kind of structure diagram of the second embodiment of user comment analytical equipment provided by the invention.
Specific embodiment
An embodiment of the present invention provides a kind of user comment analysis method and devices, solve and lack one kind in the prior art
The technical issues of method effectively analyzed to user comment.
In order to make the invention's purpose, features and advantages of the invention more obvious and easy to understand, below in conjunction with the present invention
Attached drawing in embodiment is clearly and completely described the technical solution in the embodiment of the present invention, it is clear that disclosed below
Embodiment be only part of the embodiment of the present invention, and not all embodiment.Based on the embodiments of the present invention, this field
All other embodiment that those of ordinary skill is obtained without making creative work, belongs to protection of the present invention
Range.
Inventor has found that user comment includes much words to businessman without reference to value, such as " production under study for action
This word of product ", and word frequency of this word without reference to value within each period is roughly the same, thus need by
The word for having reference value screens, therefore the method for employing word frequency, special for the ease of understanding technical scheme of the present invention
Word frequency is illustrated：
Word frequency (term frequency, TF) refers to the number that some given word occurs in this document；It is false
If total word number of a file is 100, and word " cow " occurs 3 times, then " cow " word is in this document
Word frequency is exactly 0.03 (3/100).
Referring to Fig. 1, a kind of flow diagram of the first embodiment of user comment analysis method provided by the invention.
The present invention provides a kind of first embodiment of user comment analysis method, including：
Step 101, word segmentation processing is carried out to user comment to be analyzed, every user comment is resolved into word.
Such as user comment to be analyzed is " this commodity is fine ", then the word that this user comment resolves into is " this
Money ", " commodity ", " fine " or " this ", " commodity ", " fine ", " good ", it should be noted that point of both word segmentation results
Word degree is different, so specific participle degree can be adjusted according to actual needs.
Step 102, using preset time as the period, the word frequency of each word in nearest two periods is counted respectively.
Because businessman generally thinks to analyze user comment as soon as possible and obtains valuable information, the embodiment of the present invention will
Two periods for seeking statistics word frequency are nearest two periods, and the period can be actually needed according to businessman and be configured, such as one week
Or one month.
It should be noted that the word word frequency in nearest multiple periods can also be counted, so as to pass through word word frequency
Analysis of trend word the problem of corresponding to whether obtain solution to a certain extent.
Step 103, the word that word frequency difference absolute value in two periods is more than first threshold is exported.
First threshold can be configured according to actual needs, and there are many output forms of word, such as can be table
Form or block diagram form.
Referring to Fig. 2, a kind of flow diagram of the second embodiment of user comment analysis method provided by the invention.
The present invention provides a kind of second embodiment of user comment analysis method, including：
Step 201, negative user comment is selected from all user comments crawled according to emotion, and will be negative
User comment is as user comment to be analyzed.
It is understood that before step 201 is carried out, it is necessary first to obtain user comment, acquisition modes can be logical
Reptile instrument is crossed to crawl from the user comment of application market correlation APP or webpage.
Because businessman analyzes user comment, primarily to pinpointing the problems and then solving the problems, such as, such as timely
The deficiency of itself product is solved, and then takes corrective measure, also can provide decision support to new product development etc., it is possible to from institute
Have and negative user comment is selected in user comment, then negative user comment is targetedly analyzed.
Wherein, the method for selecting negative user comment from all user comments crawled according to emotion has very
It is more, such as can specifically include：
According to the preset emotion word included in user comment weight corresponding with each preset emotion word
Negative user comment is picked out from all user comments crawled.
Preset emotion word can include " good ", " bad ", " slow " and " poor " etc..
Each the corresponding weight of preset emotion word can be set according to actual needs.
It is commented on to select negative user, and possible existing front emotion word in a user comment, also have negative
Face emotion word, such as " clothes material for making clothes is fine, and delivery is fast, and attitude is good, but clothes has aberration " such user
Comment is commented on, but have the negative reviews of remaining one side in spite of the front of three aspects, and this respect of negative reviews is independent
The problem of evaluating aberration, so need to select this user comment, the method that can be set by weight, such as by Negative sentiments
The weight of color word is set as the weight much larger than front emotion word.
Step 202, word segmentation processing is carried out to user comment to be analyzed, every user comment is resolved into word.
Step 202 is identical with the content of step 101 in the application first embodiment, and specific descriptions may refer to the first implementation
The content of example step 101, details are not described herein.
Step 203, using preset time as the period, the word frequency of each word in nearest two periods is counted respectively.
Step 203 is identical with the content of step 102 in the application first embodiment, and specific descriptions may refer to the first implementation
The content of example step 102, details are not described herein.
Step 204, the word that word frequency difference absolute value in two periods is more than first threshold is exported.
Step 204 is identical with the content of step 103 in the application first embodiment, and specific descriptions may refer to the first implementation
The content of example step 103, details are not described herein.
Step 205, classification processing is carried out to user comment to be analyzed, user comment is classified as corresponding classification, every
User comment corresponds to one or more classifications.
For example, " clothes material for making clothes is fine, and delivery is fast, and attitude is good, but clothes has aberration " such user comment was both
Quality category can be classified as, and field of express delivery can be classified as, service field can also be classified as.
Wherein, the classification processing method carried out to every user comment has very much, such as can specifically include：
According to the preset classification keyword and the corresponding weight of each classification keyword included in user comment, user is calculated
The weight of comment and, by user comment be included into second threshold be less than weight sum classification, wherein each classification correspond to one second
Threshold value, and corresponding one or more classification keywords.
Such as clothes quality category can include material for making clothes, elasticity and color these three keywords, and material for making clothes and the two passes of color
Keyword corresponds to respective weight respectively, and " clothes material for making clothes is fine, and delivery is fast, and attitude is good, but clothes has aberration " is one such
User comment includes two keywords of material for making clothes and color, and calculate the two keywords and respective weights seizes the opportunity and obtain weight
With this user comment is then classified as clothes quality this class by weight and the second threshold if more than this classification of clothes quality
Not.
In order to ensure the promptness of user comment analysis, classification processing can carry out, that is, crawl to a user and comment in real time
By just carrying out classification processing.
It should be noted that step 205 need are after step 201, there is no sequencing with step 202.
Step 206, according to the preset emotion word included in the user comment, the preset interference word included,
Each preset corresponding weight of emotion word and the length of the user comment to user comment to be analyzed into
Row emotion divides.
Such as can user comment be divided by favorable user comment, negative user comment according to preset emotion word
It is commented on intermediate user, intermediate user comment refers to dead-pan user comment, such as " product is used not yet ", may be used also
With according to each the corresponding weight of the preset emotion word comments on negative user and favorable user is commented on into traveling one
Step divides, and specific divided rank can be adjusted according to actual needs, such as negative user comment is poor and very poor respectively
Favorable user comment is also divided into good and fabulous two grades by two grades.
In emotion partition process, need to consider some interference words, it is also necessary to consider the length of user comment, example
Such as " clothes material for making clothes is fine, and delivery is fast, and attitude is good, but clothes has aberration " and " clothes has aberration " this two user comments,
" clothes material for making clothes is fine, and delivery is fast, and attitude is good, but clothes has aberration " can be divided into difference, but clothes is had aberration "
It is divided into very poor.
Referring to Fig. 3, the present invention provides a kind of structure diagrams of the first embodiment of user comment analytical equipment.
The present invention provides a kind of first embodiment of user comment analytical equipment, including：
Every user comment for carrying out word segmentation processing to user comment to be analyzed, is resolved into word by word-dividing mode 301
Language.
Word frequency statistics module 302, for using preset time as the period, counting each word in nearest two periods respectively
Word frequency.
Output module 303, for exporting the word that word frequency difference absolute value in two periods is more than first threshold.
Referring to Fig. 4, the present invention provides a kind of structure diagrams of the second embodiment of user comment analytical equipment.
The present invention provides a kind of second embodiment of user comment analytical equipment, including：
Module 401 is chosen in comment, for selecting negative user from all user comments crawled according to emotion
Comment, and negative user is commented on as user comment to be analyzed.
Comment selection module 401 selects negative user comment according to emotion from all user comments crawled can
To specifically include：
According to the preset emotion word included in user comment weight corresponding with each preset emotion word
Negative user comment is picked out from all user comments crawled.
Every user comment for carrying out word segmentation processing to user comment to be analyzed, is resolved into word by word-dividing mode 402
Language.
Word frequency statistics module 403, for using preset time as the period, counting each word in nearest two periods respectively
Word frequency.
Output module 404, for exporting the word that word frequency difference absolute value in two periods is more than first threshold.
User comment for carrying out classification processing to user comment to be analyzed, is classified as corresponding class by sort module 405
Not, every user comment corresponds to one or more classifications.
It should be noted that sort module 405 can carry out real-time grading processing.
Sort module 405 can specifically include the classification processing that every user comment carries out：
According to the preset classification keyword and the corresponding weight of each classification keyword included in user comment, user is calculated
The weight of comment and, by user comment be included into second threshold be less than weight sum classification, wherein each classification correspond to one second
Threshold value, and corresponding one or more classification keywords.
Emotion division module 406, for according to the preset emotion word included in the user comment, include
Preset interference word, each preset corresponding weight of emotion word and the user comment length treat point
The user comment of analysis carries out emotion division.
Analysis result display module 407, for that will show the analysis result of user comment.
Analysis result includes word frequency difference absolute value in two periods of output and is more than the word of first threshold, classification handling result
With emotion division result, specific exhibition method can be succinct report form.
The above, the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although with reference to before
Embodiment is stated the present invention is described in detail, it will be understood by those of ordinary skill in the art that：It still can be to preceding
The technical solution recorded in each embodiment is stated to modify or carry out equivalent replacement to which part technical characteristic；And these
Modification is replaced, the spirit and scope for various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution.
Claims (10)
1. a kind of user comment analysis method, which is characterized in that including：
Word segmentation processing is carried out to user comment to be analyzed, every user comment is resolved into word；
Using preset time as the period, the word frequency of each word in nearest two periods is counted respectively；
Export the word that word frequency difference absolute value in two periods is more than first threshold.
2. user comment analysis method according to claim 1, which is characterized in that further include：
Classification processing is carried out to the user comment to be analyzed, the user comment is classified as corresponding classification, described in every
User comment corresponds to one or more classifications.
3. user comment analysis method according to claim 2, which is characterized in that carried out to user comment every described
Classification processing specifically includes：
According to the preset classification keyword included in the user comment and the corresponding weight of each classification keyword, calculate
The weight of the user comment and, by the user comment be included into second threshold be less than the weight sum classification, wherein each
Classification corresponds to a second threshold, and corresponding one or more classification keywords.
4. user comment analysis method as claimed in any of claims 1 to 3, which is characterized in that treated to described
Before the user comment of analysis carries out word segmentation processing and classification processing, further include：
Negative user comment is selected from all user comments crawled according to emotion, and the negative user is commented on
As the user comment to be analyzed.
5. user comment analysis method according to claim 1, which is characterized in that further include：
According to the preset emotion word included in the user comment, the preset interference word included, each described preset
The corresponding weight of emotion word and the length of the user comment carry out emotion to user comment to be analyzed and draw
Point.
6. a kind of user comment analytical equipment, which is characterized in that including：
Every user comment for carrying out word segmentation processing to user comment to be analyzed, is resolved into word by word-dividing mode；
Word frequency statistics module, for using preset time as the period, counting the word frequency of each word in nearest two periods respectively；
Output module, for exporting the word that word frequency difference absolute value in two periods is more than first threshold.
7. user comment analytical equipment according to claim 6, which is characterized in that further include：
For carrying out classification processing to the user comment to be analyzed, the user comment is classified as accordingly for sort module
Classification, every user comment correspond to one or more classifications.
8. user comment analytical equipment according to claim 7, which is characterized in that the sort module is to use every described
The classification processing that family comment carries out specifically includes：
According to the preset classification keyword included in the user comment and the corresponding weight of each classification keyword, calculate
The weight of the user comment and, by the user comment be included into second threshold be less than the weight sum classification, wherein each
Classification corresponds to a second threshold, and corresponding one or more classification keywords.
9. the user comment analytical equipment according to any one in claim 6 to 8, which is characterized in that further include：
Module is chosen in comment, for negative user comment to be selected from all user comments crawled according to emotion, and
The negative user is commented on as the user comment to be analyzed.
10. user comment separating device according to claim 9, which is characterized in that further include：Emotion division module,
For according to the preset emotion word included in the user comment, the preset interference word included, each described preset
The corresponding weight of emotion word and the length of the user comment carry out emotion to user comment to be analyzed and draw
Point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711484686.6A CN108228563A (en) | 2017-12-29 | 2017-12-29 | A kind of user comment analysis method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711484686.6A CN108228563A (en) | 2017-12-29 | 2017-12-29 | A kind of user comment analysis method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108228563A true CN108228563A (en) | 2018-06-29 |
Family
ID=62646364
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711484686.6A Pending CN108228563A (en) | 2017-12-29 | 2017-12-29 | A kind of user comment analysis method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108228563A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109800417A (en) * | 2018-12-14 | 2019-05-24 | 平安普惠企业管理有限公司 | Analysis method, device, equipment and the computer storage medium of application software experience |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100049499A1 (en) * | 2006-11-22 | 2010-02-25 | Haruo Hayashi | Document analyzing apparatus and method thereof |
| CN103823792A (en) * | 2014-03-07 | 2014-05-28 | 网易（杭州）网络有限公司 | Method and equipment for detecting hotspot events from text document |
| CN105956031A (en) * | 2016-04-25 | 2016-09-21 | 深圳市永兴元科技有限公司 | Text classification method and apparatus |
| CN106294425A (en) * | 2015-05-26 | 2017-01-04 | 富泰华工业（深圳）有限公司 | The automatic image-text method of abstracting of commodity network of relation article and system |
| CN107506424A (en) * | 2017-08-17 | 2017-12-22 | 国网北京市电力公司 | Data analysing method and device |
-
2017
- 2017-12-29 CN CN201711484686.6A patent/CN108228563A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100049499A1 (en) * | 2006-11-22 | 2010-02-25 | Haruo Hayashi | Document analyzing apparatus and method thereof |
| CN103823792A (en) * | 2014-03-07 | 2014-05-28 | 网易（杭州）网络有限公司 | Method and equipment for detecting hotspot events from text document |
| CN106294425A (en) * | 2015-05-26 | 2017-01-04 | 富泰华工业（深圳）有限公司 | The automatic image-text method of abstracting of commodity network of relation article and system |
| CN105956031A (en) * | 2016-04-25 | 2016-09-21 | 深圳市永兴元科技有限公司 | Text classification method and apparatus |
| CN107506424A (en) * | 2017-08-17 | 2017-12-22 | 国网北京市电力公司 | Data analysing method and device |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109800417A (en) * | 2018-12-14 | 2019-05-24 | 平安普惠企业管理有限公司 | Analysis method, device, equipment and the computer storage medium of application software experience |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106485562B (en) | Commodity information recommendation method and system based on user historical behaviors | |
| Maertens et al. | Horticultural exports, female wage employment and primary school enrolment: Theory and evidence from Senegal | |
| Pulaj et al. | The impact of generic competitive strategies on organizational performance. The evidence from Albanian context | |
| US20150248412A1 (en) | Information pushing method and apparatus | |
| CN108132964A (en) | A kind of collaborative filtering method to be scored based on user item class | |
| TW201224972A (en) | Sorting method and apparatus of query results | |
| CN107688988A (en) | A kind of financial product real-time recommendation method | |
| Kosasih et al. | The influence of ambidextrous organization and authentic followership on innovative performance: The mediating role of change readiness | |
| CN103942604B (en) | Forecasting Methodology and system based on forest discrimination model | |
| CN104572775A (en) | Advertisement classification method, device and server | |
| CN108228563A (en) | A kind of user comment analysis method and device | |
| CN111125561A (en) | Network heat display method and device | |
| CN109711484A (en) | A kind of classification method and system of customer | |
| CN110046951A (en) | A kind of trading activity judgment method and system | |
| Miller et al. | The motivation for foreign direct investment | |
| Radovanović et al. | Behaviour of Millenial wine consumers in southern Serbia | |
| CN110147965A (en) | A kind of customer analysis system based on big data | |
| CN109242610A (en) | A kind of books purchasing management system based on mobile payment | |
| CN105868991A (en) | Method and device for identifying machine assisted cheating | |
| CN108241989A (en) | Otherness data capture method and device | |
| Shil et al. | Factors influencing the choice of sophisticated management accounting practices-exploratory evidence from an emerging market | |
| Ciucescu et al. | The performance of Non-Governmental Organizations | |
| Muang et al. | Customer Loyalty: Improve customer relationship management that impact the customer satisfaction | |
| Lee et al. | Does satisfaction make customers repurchase same brand again? | |
| CN110135467A (en) | A kind of model training method, device, system and recording medium based on data splicing |
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: 20180629 |