CN108549727A - User's profit information-pushing method based on web crawlers and big data analysis - Google Patents
User's profit information-pushing method based on web crawlers and big data analysis Download PDFInfo
- Publication number
- CN108549727A CN108549727A CN201810409419.0A CN201810409419A CN108549727A CN 108549727 A CN108549727 A CN 108549727A CN 201810409419 A CN201810409419 A CN 201810409419A CN 108549727 A CN108549727 A CN 108549727A
- Authority
- CN
- China
- Prior art keywords
- user
- profit
- information
- profit information
- condition
- 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
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
It can be realized to qualified user push profit information to provide one kind, it can ensure the comprehensive user's profit information-pushing method of information again, the present invention provides a kind of user's profit information-pushing method based on web crawlers and big data analysis, includes the following steps：Step S1 obtains the user information of user；Step S2 carries out text analyzing to user information and obtains reflection condition flag；Step S3 classifies user according to condition flag to obtain different user's classification；Step S4 obtains profit information using web crawlers from multiple information issuing web sites；Step S5 carries out text analyzing to profit information and obtains the condition of acquiring profit corresponding to different profit information；Step S6, according to condition flag and condition of acquiring profit obtain user's classification and the matching degree of profit information；Step S7, the user in being classified to user according to the matching degree between profit information push profit information.
Description
Technical field
The present invention relates to a kind of information-pushing methods, and in particular to a kind of use based on web crawlers and big data analysis
Family profit information-pushing method.
Background technology
Modern society is the society of information explosion, and how from bulk information obtain with the relevant information of number one from
And it is then that current Internet user pays close attention to more problem to make a profit, for example, in order to promote the volume of the flow of passengers, eating and drinking establishment is often in difference
Website on publication meet specified conditions (such as holding college student card or certain bank card etc.) can be in limiting time
The information of discount taken can be obtained discount when user knows that the information and self-condition meet；In another example in order to promote
Development in science and technology, government organs would generally provide with funds to qualified scientific and technological enterprises or policy in terms of support etc..
However, since the above-mentioned relevant profit information of interests obtained from user usually will disperse publication in different websites
It is upper that (for example, local information issuing web site, bank's favor information issues the page or the bulletin page etc. of different government organs website
Deng), user almost can not independently, comprehensively obtain these information.Number of site can use householder after screening to condition coupling
Dynamic pushed information, but the range of this push is only limitted to log in and browsed the user of the website, and other users can not then obtain
It pushes and cannot learn the profit information on the website.
In order to promote the comprehensive of information, number of site, which additionally uses, obtains the profit information of other multiple website orientations simultaneously
The mode being shown so that user can browse through and find useful profit information from these information.However, such side
It is excessive that formula results in information content again, and user is caused to have to take a significant amount of time and browse and screen.Further, since such
Profit information content is huge in website, user conditional information is also extremely complex, therefore screens and condition coupling mechanism has also been difficult to
Effect is realized.
Invention content
Profit information is pushed to qualified user and is protected to solve the above problems, providing one kind and can realize
The comprehensive user's profit information-pushing method of information is demonstrate,proved, present invention employs following technical solutions：
The present invention provides a kind of user's profit information-pushing method based on web crawlers and big data analysis, is used for
Corresponding profit information is pushed to different users, which is characterized in that include the following steps to according to user information：Step S1,
Obtain the user information of user；Step S2 carries out text analyzing to user information successively, respectively obtains and reflects obtaining for each user
Multiple condition flags of sharp condition；Step S3 classifies to user according to condition flag, to obtain different user point
Class；Step S4, obtained from multiple information issuing web sites associated with user's profit using web crawlers with make a profit it is relevant
Profit information；Step S5 carries out text analyzing to profit information, obtains the condition of acquiring profit corresponding to different profit information；Step
Rapid S6, according to and user classify corresponding condition flag and condition of acquiring profit corresponding with profit information obtain user's classification with
The matching degree of profit information；Step S7, the user in being classified to user according to the matching degree between profit information push
Profit information.
User's profit information-pushing method provided by the invention based on web crawlers and big data analysis, can also have
There is such technical characteristic, wherein step S3 includes the following steps：Step S3-1 is formed condition flag corresponding with user
Feature vector；Step S3-2, feature based vector cluster user；Step S3-3 obtains the use per class user successively
Family label.
User's profit information-pushing method provided by the invention based on web crawlers and big data analysis, can also have
There is such technical characteristic, wherein the cluster of step S3-2 is carried out using the clustering algorithm based on community discovery.
User's profit information-pushing method provided by the invention based on web crawlers and big data analysis, can also have
There is such technical characteristic, wherein step S5 includes the following steps：Step S5-1, to profit information carry out data cleansing to
Remove invalid information therein；Step S5-2 carries out text analyzing to the profit information after cleaning, obtains each profit information institute
Including condition of acquiring profit.
User's profit information-pushing method provided by the invention based on web crawlers and big data analysis, can also have
There is such technical characteristic, wherein in step S6, matching degree is condition satisfaction, and step S6 includes the following steps：Step
S6-1, judges whether each condition flag of enterprise classifying is consistent with the condition of acquiring profit in profit information, is given if meeting
One high matching value gives a low matching value if not meeting；Step S6-2 sets different types of condition of acquiring profit different
Weighted value；Matching value is multiplied by after corresponding weighted value and sums up calculating by step S6-3, obtains enterprise classifying and letter of making a profit
Condition satisfaction between breath.
User's profit information-pushing method provided by the invention based on web crawlers and big data analysis, can also have
There is such technical characteristic, wherein the push of step S7 is carried out according to following rule：Matching degree threshold value is set, as user point
It then sends and is higher than to the lower user of the user classification when matching degree between class and profit information is higher than matching degree threshold value
Profit information with degree threshold value.
Invention effect
According to user's profit information-pushing method provided in this embodiment based on web crawlers and big data analysis, by
In carrying out text analyzing to user information to obtain multiple condition flags, be classified to user according to condition flag,
Text analyzing also has been carried out to profit information and has obtained multiple condition of acquiring profit simultaneously, therefore can be according to condition flag and profit
Condition obtains the matching degree of user's classification and profit information, to which each user in classifying to user pushes profit information,
It allows users to obtain the profit information more met with self-condition, allows user that can either obtain profit information, and do not have to
It devotes a tremendous amount of time and carries out screening and condition coupling.
Moreover, since in the above process, user has been divided into different classification, and condition flag is classified with user
It is corresponding, therefore the condition coupling carried out in the process is to be carried out based on classification rather than carried out based on single user.Institute
With even if more than based on each user if needing to calculate each classification and its calculation amount of the condition satisfaction of profit information one by one
The matching of progress wants small, can be completed quickly and effectively condition coupling.
Description of the drawings
Fig. 1 is user's profit information-pushing method based on web crawlers and big data analysis of the embodiment of the present invention
Flow chart.
Specific implementation mode
With reference to the accompanying drawings and embodiments come illustrate the present invention specific implementation mode.In embodiment, need to obtain with enterprise
Government organs publication fund policy support information scene for illustrate, that is, in embodiment, these enterprises as use
Family has the demand of the profits information such as the policy support information for obtaining and being consistent with the condition of its own.
<Embodiment>
Fig. 1 is user's profit information-pushing method based on web crawlers and big data analysis of the embodiment of the present invention
Flow chart.
As shown in Figure 1, user's profit information-pushing method based on web crawlers and big data analysis of the present embodiment
Include the following steps.
Step S1 obtains the user information of user.
In the present embodiment, user is enterprise, and user information includes that the enterprises such as title, address, the contact method of user are basic
Information and business license information, shareholder and investment information, key personnel's information, branch's information, clearance information, administration are permitted
Can information, administrative penalty information, whether be included in the abnormal register of operation and whether be included in the enterprises such as enterprise's list of breaking one's promise that break the law on a serious scale
Operation information.Wherein, enterprise's essential information can be provided by enterprise in registration, and enterprise operation information can be carried by enterprise in registration
For that can also be obtained from related public notification of information website in the case where obtaining enterprise's license.
Step S2 carries out text analyzing to user information successively, respectively obtains the more of the condition of acquiring profit for reflecting each user
A condition flag.Wherein, every enterprise operation information includes characteristic attribute and characteristic value.
In the present embodiment, when enterprise operation information obtained from public notification of information website is a plurality of text, to these texts point
Not carry out text analyzing can be obtained condition flag.For example, using the dictionary being previously stored to " first False Company Registration Capital 3000
Ten thousand " this text are segmented, and the registered capital further according to the context relationship judgement first company of text is " 30,000,000 "；This
When, " 30,000,000 " are exactly a condition flag, and characteristic attribute is " registered capital ", and characteristic value is " 30,000,000 ".
In addition, when enterprise operation information is provided by enterprise, allow user in specified column fill substance due to may be used
Mode (for example, user is directly allowed to fill in content as " 30,000,000 " in " registered capital " this column), therefore can not
The characteristic attribute and characteristic value of condition flag can be obtained with carrying out the text analyzing such as segmenting.
After acquisition being analyzed to the condition flag of whole users, you can enter step S3.
Step S3 classifies to user according to condition flag, to obtain different user's classification.
In the present embodiment, the user's portrait treatment technology that is classified based on of user is carried out, is included the following steps：
Condition flag is formed feature vector corresponding with user by step S3-1.Due to can obtain policy support with
The qualification and business circumstance of enterprise are relevant, and the condition flag in the present embodiment can reflect qualification and the operation of enterprise
Situation, therefore whole condition flags is used to form feature vector；In other embodiments, different applications can also be directed to
The required condition of scene, selects different condition flags to form feature vector.
Step S3-2, feature based vector cluster user, that is, using the clustering algorithm of feature based vector to enterprise
Industry is clustered.In the present embodiment, enterprise is clustered using the clustering algorithm in the prior art based on community discovery, from
And obtain different enterprise classifying (i.e. user classifies).In addition, in other embodiments, can also use it is in the prior art its
His clustering algorithm, such as K-means clustering algorithms, hierarchical clustering algorithm are clustered.
In the present embodiment, the concrete methods of realizing of the clustering algorithm based on community discovery is：To all users (i.e. enterprise)
Related coefficient is calculated according to feature vector, so obtains the correlation matrix between all users.When phase between two users
When relationship number is more than preset threshold value, then by the two user Lian Bian, can so it obtain using all users as node
Network.User is equivalent to each point, the structure of whole network is constituted between user by the frontier juncture system that is mutually connected, such
In network, the connection between some users is more close, and the connection relation between some users is more sparse, and connection is more close
Part can be seen as a community, have more close connection between internal node, and a then phase between the communities Liang Ge
More sparse to connecting, this is just known as community structure.Based on this, community's hair is carried out using existing fast unfolding algorithms
Existing, the set comprising node is user group in obtained each community, that is, realizes the cluster of user.
After cluster, certain similitude, such as business similitude, registered place similitude, registration are all had per class enterprise
Capital similitude etc..That is, it is more similar per class enterprise, and in one aspect or certain several respects similitude is higher.
Step S3-3 obtains the user tag per class user successively.
As described above, after being clustered to enterprise according to above-mentioned clustering algorithm, then all had per class enterprise certain similar
Property.Therefore, by existing user portrait treatment technology, according to common feature possessed by every class enterprise (for example, common item
Part feature) obtain multiple user tags of all kinds of enterprises.At this point, each user tag represents one of the enterprise of the category
Common trait, such as be co-registered with etc..
Step S4 is obtained and profit phase using web crawlers from multiple information issuing web sites associated with user's profit
The profit information of pass.
In the present embodiment, information issuing web site associated with the profit of the enterprise as user is mainly various policy branch
Information issuing web site and policy news issuing web site are held, for example, the preferential tax revenue issued for the enterprise of certain poverty-stricken area
Policy information may be issued on the bulletin page of district government website, and talent introduction supports information that may issue in the regional talent
On the bulletin page of center website, these information are to be published to the public by open channel, therefore the present embodiment is used and climbed
Worm technology is obtained and is stored one by one to these information.
Step S5 carries out text analyzing to profit information, obtains the condition of acquiring profit corresponding to different profit information, specifically
Include the following steps：
Step S5-1 carries out data cleansing to profit information, removes invalid information.Since the concrete form of profit information is logical
Often it is bulletin or news, therefore will contains redundancy (such as website largely unrelated with profit after the crawl of reptile module
Network address, webpage parameter etc.), these redundancies can be removed after data cleansing.Data cleansing may be used based on just
Then the mode of expression formula carries out, to which removal meets the information of redundancy form (such as meeting the textual form of network address).
Step S5-2 carries out text analyzing to the profit information after cleaning, obtains the profit that each profit information is included
Condition.The profits information such as bulletin news form different articles after cleaning, by segmenting, extracting the texts such as keyword point
The mode of analysis can be obtained the specific condition of acquiring profit that each item profit information is included.
In the present embodiment, participle and extraction keyword can be carried out based on the dictionary being previously stored, meanwhile, each of acquisition
Condition of acquiring profit contains multiple profit keywords with different attribute.For example, to " proposing to A to providing 15 or more
The university students' innovative undertaking enterprise of job provides foundation guaranteed loan government discount ... ... " segmented, extract keyword after,
Obtain " A " (registered place), " 15 or more " (job number), " university students' innovative undertaking enterprise " (enterprise nature), " guarantee of starting an undertaking
The profits keyword such as loan " (profit project), " government's discount " (profit content), wherein the content in bracket is key of making a profit
The corresponding attribute of word.In addition, the present embodiment is also by a pre-set replacement dictionary come the identical but word to some meanings
Different profit keywords is replaced and uniformly, causes substantially identical profit to be closed because selecting vocabulary different to avoid appearance
Keyword is mistaken as different situations.
By above-mentioned data scrubbing and text analyzing, each condition of acquiring profit just is extracted out respectively from profit information,
These condition of acquiring profit can be stored together with its acquisition time, subsequently into step S6.
Step S6 is obtained according to and user's corresponding condition flag of classification and condition of acquiring profit corresponding with profit information
The matching degree that user classifies with profit information.
In the present embodiment, user, which classifies, to be indicated with the matching degree of profit information using condition satisfaction.
Condition satisfaction refers to the degree for the condition of acquiring profit that enterprise meets in profit information, uses matching value and weighted value knot
The mode of conjunction obtains to calculate, and specific computation rule is as follows：
(sentence first, it is determined that whether each condition flag of enterprise classifying meets each condition of acquiring profit in profit information
Whether the disconnected identical condition flag of attribute consistent with condition of acquiring profit), a high matching value is given if meeting, if do not meet to
It gives a low matching value and matching value is even denoted as 0.
Then, the influence according to different types of condition to satisfaction degree sets weight to different types of condition of acquiring profit
Value, matching value is multiplied by after corresponding weighted value and sums up calculating, you can obtains some enterprise classifying and some profit information
Between condition satisfaction.
For example, ought include in the condition flag of an enterprise classifying " A " (registered place), while in a profit information
When including the condition of acquiring profit of " A " (registered place), since the two is consistent, given for " registered place " this condition
One high matching value；Simultaneously as enterprise manage support policy in terms of, registered place be very important condition (if for example,
Registered place can not possibly then not enjoy its support policy completely in somewhere), therefore give a high weighted value for this condition.Separately
Outside, in above-mentioned calculating process, high matching can only be set with the matching value of the condition flag of written form storage and condition of acquiring profit
Value and two grades of low matching value, but what is stored in digital form can then set high, normal, basic equal different brackets to embody condition spy
The degree of closeness of sign and condition of acquiring profit.
Above-mentioned calculating carries out in the way of traversing one by one, i.e., each enterprise classifying is carried out with each profit information successively
The calculating with condition satisfaction is compared, the condition satisfaction between different enterprise classifyings and all profit information is obtained, then may be used
Enter step S7.
Step S7, the user in being classified to user according to the matching degree between profit information push profit information.
After the matching degree (i.e. condition satisfaction) for obtaining different enterprise classifyings and all profit information by step S6, i.e.,
The profit information of each enterprise transmission matching degree high (i.e. condition satisfaction numerical value is high) that can be into enterprise classifying.
The profit information that needs are sent is determined in the present embodiment by the way of according to condition satisfaction threshold value.For example,
Condition satisfaction between some enterprise classifying and several profit information is above preset threshold value, then under the classification
All enterprises send these profit information for being higher than threshold value (contact method provided when can be according to enterprises registration is sent).
After enterprise receives corresponding profit information, you can judge whether itself can therefrom obtain according to bulletin therein or news content
Profit simultaneously makes Corresponding Countermeasures.
In the above process, since the present embodiment determines the mode of profit information for needing to send using based on threshold value,
Therefore enterprise probably receives condition satisfaction and is higher than threshold value but actually self-condition and the profit non-fully strictly met
Information.In this case, although enterprise also not exclusively has corresponding condition, but still can be according to profit information in shortcoming
Aspect is made an effort, and is supported to obtain corresponding fund policy afterwards.
Embodiment effect
According to user's profit information-pushing method provided in this embodiment based on web crawlers and big data analysis, by
In carrying out text analyzing to user information to obtain multiple condition flags, be classified to user according to condition flag,
Text analyzing also has been carried out to profit information and has obtained multiple condition of acquiring profit simultaneously, therefore can be according to condition flag and profit
Condition obtains the matching degree of user's classification and profit information, to which each user in classifying to user pushes profit information,
It allows users to obtain the profit information more met with self-condition, allows user that can either obtain profit information, and do not have to
It devotes a tremendous amount of time and carries out screening and condition coupling.
Moreover, since in the above process, user has been divided into different classification, and condition flag is classified with user
It is corresponding, therefore the condition coupling carried out in the process is to be carried out based on classification rather than carried out based on single user.Institute
With even if more than based on each user if needing to calculate each classification and its calculation amount of the condition satisfaction of profit information one by one
The matching of progress wants small, can be completed quickly and effectively condition coupling.
In embodiment, as a result of the condition satisfaction obtained based on matching value and weighted value come weigh user classification with
Matching degree between profit information, thus can not only reflect on the whole user classification between different profit information
Matching degree, additionally it is possible to so that important condition is generated bigger to the result of matching degree influences, to keep matching degree more accurate
Really, but also the push of subsequent information is more accurate.
In addition, the profit information push of the present embodiment is carried out according to preset matching degree threshold value, therefore can
Also it is pushed when user not exclusively meets the condition of acquiring profit of profit information, user is allowed more to be possible to meet
Profit information simultaneously allows user that can determine following developing direction according to these profit information, to generate more permanent promotion
Effect, is conducive to the long term growth of user.
In embodiment, to further including that profit keyword is replaced and is unified during the text analyzing of profit information
Step, therefore can avoid the occurrence of causes substantially identical profit keyword to be mistaken as different feelings because selecting vocabulary different
Condition.
Above-described embodiment is merely to illustrate the specific implementation mode of the present invention, and the present invention based on web crawlers and big
User's profit information-pushing method of data analysis is not limited to the above embodiments the range of description.
For example, in embodiment, matching degree is indicated using the condition satisfaction obtained based on matching value and weighted value
's.But in the present invention, matching degree can also be directly based upon matching value and obtain, or be directly based upon condition flag and profit item
The number that part is consistent show that such simplified way cannot reflect matching degree on the whole, but it is more simple at some
It can be applied in application scenarios single, that more globalities considerations need not be have ever made, and the method application of the present invention can be reduced
Calculation amount when these scenes, to improve efficiency.
In embodiment, the push of profit information is carried out according to preset matching degree threshold value, thus user can be allowed to obtain
The profit information of condition can't be fully met at present by obtaining some.But user in the present invention, can also be allowed to be selected in registration
Whether need to obtain such profit information, when user's selection does not need, just by the condition flag of the user and matching degree
The obtaining corresponding to the profit information of (that is, the matching degree classified with the user where the user is more than threshold value) more than threshold value
Sharp condition is compared one by one, and when exactly matching just profit information is sent so that user can according to oneself
Wish selection only obtains the profit information for the condition that complies fully with, improves information acquisition efficiency.
Claims (6)
1. a kind of user's profit information-pushing method based on web crawlers and big data analysis, for according to user information
Corresponding profit information is pushed to different users, which is characterized in that is included the following steps：
Step S1 obtains the user information of the user；
Step S2 carries out text analyzing to the user information successively, respectively obtains the condition of acquiring profit for reflecting each user
Multiple condition flags；
Step S3 classifies to the user according to the condition flag, to obtain different user's classification；
Step S4, obtained from multiple information issuing web sites associated with user's profit using web crawlers with make a profit it is relevant
Profit information；
Step S5 carries out text analyzing to the profit information, obtains the condition of acquiring profit corresponding to different profit information；
Step S6, according to and the user classify the corresponding condition flag and corresponding with the profit information described
Condition of acquiring profit obtains the matching degree of the user classification and the profit information；
Step S7, the user in being classified to the user according to the matching degree between the profit information push
The profit information.
2. user's profit information-pushing method according to claim 1 based on web crawlers and big data analysis,
It is characterized in that：
Wherein, step S3 includes the following steps：
The condition flag is formed feature vector corresponding with the user by step S3-1；
Step S3-2 clusters the user based on described eigenvector；
Step S3-3 obtains the user tag of user described in every class successively.
3. user's profit information-pushing method according to claim 1 based on web crawlers and big data analysis,
It is characterized in that：
Wherein, the cluster of step S3-2 is carried out using the clustering algorithm based on community discovery.
4. user's profit information-pushing method according to claim 1 based on web crawlers and big data analysis,
It is characterized in that：
Wherein, step S5 includes the following steps：
Step S5-1 carries out data cleansing to remove invalid information therein to the profit information；
Step S5-2 carries out text analyzing to the profit information after cleaning, show that each profit information is included
Condition of acquiring profit.
5. user's profit information-pushing method according to claim 1 based on web crawlers and big data analysis,
It is characterized in that：
Wherein, in step S6, the matching degree is condition satisfaction, and step S6 includes the following steps：
Step S6-1, judge the enterprise classifying each condition flag whether with the profit in the profit information
Condition is consistent, and a high matching value is given if meeting, and a low matching value is given if not meeting；
Step S6-2 sets different weighted values to different types of condition of acquiring profit；
Step S6-3 sums up calculating after the matching value is multiplied by the corresponding weighted value, obtains the enterprise classifying
The condition satisfaction between the profit information.
6. user's profit information-pushing method according to claim 1 based on web crawlers and big data analysis,
It is characterized in that：
Wherein, the push of step S7 is carried out according to following rule：
Matching degree threshold value is set, when the matching degree between user classification and the profit information is higher than described
The user under then classifying to the user when with degree threshold value sends the profit information higher than the matching degree threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810409419.0A CN108549727B (en) | 2018-05-02 | 2018-05-02 | User profit information pushing method based on web crawler and big data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810409419.0A CN108549727B (en) | 2018-05-02 | 2018-05-02 | User profit information pushing method based on web crawler and big data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108549727A true CN108549727A (en) | 2018-09-18 |
| CN108549727B CN108549727B (en) | 2021-11-23 |
Family
ID=63513225
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810409419.0A Active CN108549727B (en) | 2018-05-02 | 2018-05-02 | User profit information pushing method based on web crawler and big data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108549727B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109359194A (en) * | 2018-09-30 | 2019-02-19 | 北京字节跳动网络技术有限公司 | Method and apparatus for predictive information classification |
| CN111581522A (en) * | 2020-06-05 | 2020-08-25 | 预见你情感(北京)教育咨询有限公司 | Social analysis method based on user identity identification |
| CN112383544A (en) * | 2020-11-13 | 2021-02-19 | 西安热工研究院有限公司 | Service behavior portrait-based crawler resisting method suitable for electric power SCADA |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101957834A (en) * | 2010-08-12 | 2011-01-26 | 百度在线网络技术（北京）有限公司 | Content recommending method and device based on user characteristics |
| US20120066072A1 (en) * | 2007-05-25 | 2012-03-15 | Issar Amit Kanigsberg | Recommendation Systems and Methods Using Interest Correlation |
| CN102609860A (en) * | 2012-01-20 | 2012-07-25 | 彭立发 | Method and system suitable for categorizing and recommending e-commerce commodities and information |
| US20150100422A1 (en) * | 2013-10-08 | 2015-04-09 | Match.Com, L.L.C. | System and method for promoting a profile of a particular user in a network environment |
| CN105512331A (en) * | 2015-12-28 | 2016-04-20 | 海信集团有限公司 | Video recommending method and device |
| CN106446060A (en) * | 2016-09-06 | 2017-02-22 | 北京易游华成科技有限公司 | Information push and search device, method and system |
| CN107016026A (en) * | 2016-11-11 | 2017-08-04 | 阿里巴巴集团控股有限公司 | A kind of user tag determination, information-pushing method and equipment |
-
2018
- 2018-05-02 CN CN201810409419.0A patent/CN108549727B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120066072A1 (en) * | 2007-05-25 | 2012-03-15 | Issar Amit Kanigsberg | Recommendation Systems and Methods Using Interest Correlation |
| CN101957834A (en) * | 2010-08-12 | 2011-01-26 | 百度在线网络技术（北京）有限公司 | Content recommending method and device based on user characteristics |
| CN102609860A (en) * | 2012-01-20 | 2012-07-25 | 彭立发 | Method and system suitable for categorizing and recommending e-commerce commodities and information |
| US20150100422A1 (en) * | 2013-10-08 | 2015-04-09 | Match.Com, L.L.C. | System and method for promoting a profile of a particular user in a network environment |
| CN105512331A (en) * | 2015-12-28 | 2016-04-20 | 海信集团有限公司 | Video recommending method and device |
| CN106446060A (en) * | 2016-09-06 | 2017-02-22 | 北京易游华成科技有限公司 | Information push and search device, method and system |
| CN107016026A (en) * | 2016-11-11 | 2017-08-04 | 阿里巴巴集团控股有限公司 | A kind of user tag determination, information-pushing method and equipment |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109359194A (en) * | 2018-09-30 | 2019-02-19 | 北京字节跳动网络技术有限公司 | Method and apparatus for predictive information classification |
| CN109359194B (en) * | 2018-09-30 | 2020-10-20 | 北京字节跳动网络技术有限公司 | Method and apparatus for predicting information categories |
| CN111581522A (en) * | 2020-06-05 | 2020-08-25 | 预见你情感(北京)教育咨询有限公司 | Social analysis method based on user identity identification |
| CN112383544A (en) * | 2020-11-13 | 2021-02-19 | 西安热工研究院有限公司 | Service behavior portrait-based crawler resisting method suitable for electric power SCADA |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108549727B (en) | 2021-11-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Cao et al. | A cross-media public sentiment analysis system for microblog | |
| US10997244B2 (en) | Method and system for identifying and discovering relationships between disparate datasets from multiple sources | |
| Ratkiewicz et al. | Truthy: mapping the spread of astroturf in microblog streams | |
| US11048712B2 (en) | Real-time and adaptive data mining | |
| CN102208992B (en) | The malicious information filtering system of Internet and method thereof | |
| US8868609B2 (en) | Tagging method and apparatus based on structured data set | |
| CN108965245A (en) | Detection method for phishing site and system based on the more disaggregated models of adaptive isomery | |
| US20100211551A1 (en) | Method, system, and computer readable recording medium for filtering obscene contents | |
| Alomari et al. | Road traffic event detection using twitter data, machine learning, and apache spark | |
| Alomari et al. | Sentiment analysis of Arabic tweets for road traffic congestion and event detection | |
| CN107305551A (en) | The method and apparatus of pushed information | |
| US10956522B1 (en) | Regular expression generation and screening of textual items | |
| JP2011154668A (en) | Method for recommending the most appropriate information in real time by properly recognizing main idea of web page and preference of user | |
| Fayaz et al. | Machine learning for fake news classification with optimal feature selection | |
| CN103577404A (en) | Microblog-oriented discovery method for new emergencies | |
| CN107526718A (en) | Method and apparatus for generating text | |
| CN108549727A (en) | User's profit information-pushing method based on web crawlers and big data analysis | |
| CN108520007A (en) | Web page information extracting method, storage medium and computer equipment | |
| Alshehri et al. | Think before your click: Data and models for adult content in arabic twitter | |
| Kocayusufoglu et al. | Riser: Learning better representations for richly structured emails | |
| CN109857869A (en) | A kind of hot topic prediction technique based on Ap increment cluster and network primitive | |
| CN110990587B (en) | Enterprise relation discovery method and system based on topic model | |
| CN114915468A (en) | Intelligent analysis and detection method for network crime based on knowledge graph | |
| Guo et al. | Web page classification based on graph neural network | |
| Li et al. | Words analysis of online Chinese news headlines about trending events: a complex network perspective |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |