CN103617230A - Method and system for advertisement recommendation based microblog - Google Patents
Method and system for advertisement recommendation based microblog Download PDFInfo
- Publication number
- CN103617230A CN103617230A CN201310608335.7A CN201310608335A CN103617230A CN 103617230 A CN103617230 A CN 103617230A CN 201310608335 A CN201310608335 A CN 201310608335A CN 103617230 A CN103617230 A CN 103617230A
- Authority
- CN
- China
- Prior art keywords
- lexical item
- microblogging
- data
- feature
- dictionary
- 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
- 238000000034 method Methods 0.000 title claims abstract description 24
- 239000013598 vector Substances 0.000 claims abstract description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000007635 classification algorithm Methods 0.000 claims description 3
- 230000002068 genetic effect Effects 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 8
- 238000007418 data mining Methods 0.000 abstract description 2
- 238000004364 calculation method Methods 0.000 abstract 1
- 238000013507 mapping Methods 0.000 abstract 1
- 238000004458 analytical method Methods 0.000 description 6
- 230000010365 information processing Effects 0.000 description 4
- 244000097202 Rathbunia alamosensis Species 0.000 description 3
- 235000009776 Rathbunia alamosensis Nutrition 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000003796 beauty Effects 0.000 description 2
- 238000007493 shaping process Methods 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 239000002537 cosmetic Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
- G06F16/353—Clustering; Classification into predefined classes
Abstract
The invention belongs to the field of data mining and provides a method and system for advertisement recommendation based a microblog. The method comprises the steps that microblog data are read; the microblog data are initialized and a microblog text lexical item set is obtained; stop words of the microblog text lexical item set are deleted and a microblog text original feature lexical item set is obtained; mapping is conducted on the microblog text original feature lexical item set and a feature lexical item dictionary, whether lexical items in the microblog text original feature lexical item set exist in the feature lexical item dictionary or not is judged, and the tf-idf values of the appearing lexical items are calculated and serve as the feature values of the lexical items; whether the lexical items of the feature lexical item dictionary exist in the microblog text original feature lexical item set or not is judged and the feature values of the lexical items which do not appear are marked to be zero; feature vectors of the feature values obtained through calculation are automatically classified to classifications divided in advance; according to an automatic classification result, advertisements are recommended to a user. The advertisements recommended by the method and system are accurate and the effect is good.
Description
Technical field
The invention belongs to Data Mining, relate in particular to a kind of advertisement recommend method and system based on microblogging.
Background technology
Along with at home popular of the social network sites such as Sina's microblogging, Tengxun's microblogging, the social medias such as microblogging not only become the platform that netizen issues, shares, diffuses information, and have accumulated extensive netizen's behavioral data.In May, 2012, the deputy general manager Lu Yi of microblogging division department of Sina points out, the microblogging registered user of Sina surpasses 300,000,000, and average issue every day of user surpasses 100,000,000 microblogging contents.The radix of microblog users is large, data volume is large, if existing mass data can be analyzed and excavate to microblogging operation system, can to the interest of microblog users, judge comparatively accurately according to analysis result, according to the interest of microblog users, it is carried out to advertisement putting, the advertisement microblog users being pushed will make microblog users, businessman and the tripartite of microblogging operator all be benefited.
Existing microblogging advertisement recommend method mainly utilizes label in individual subscriber data or user's searching record to carry out interest judgement to microblog users, so it is pushed to user may interested advertisement.Due to a lot of users' personal information the inside, not contain the label that label or user fill in when creating personal information inaccurate, therefore by user tag, it carried out to advertisement and recommend to reach good effect.And by the searching record of microblog users being judged to user's interest has certain limitation, only can represent this user's current needs and can not judge comparatively accurately its interest.
Summary of the invention
The embodiment of the present invention provides a kind of advertisement recommend method based on microblogging, is intended to solve existing method accuracy when digging user information low, thereby causes the bad problem of advertisement recommendation effect.
The embodiment of the present invention is achieved in that a kind of advertisement recommend method based on microblogging, and described method comprises the steps:
Read user's microblogging data;
The microblogging data that initialization is read, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read;
Delete the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
The set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and the word frequency-reverse file frequency tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, using as described in appear at described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary of generation in advance in the eigenwert of microblogging;
Whether the lexical item of the feature lexical item dictionary generating in advance described in judgement appears in the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0;
Use the disaggregated model obtain in advance by user's microblogging data automatic classification in the classification of dividing in advance;
The result of automatic classification of take is foundation, to the user's recommended advertisements that reads microblogging data.
Another object of the embodiment of the present invention is to provide a kind of advertisement commending system based on microblogging, and described system comprises:
The first data are read in module, for reading user's microblogging data;
The first data initialization module, the microblogging data that read for initialization, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read;
First Characteristic extraction module, for deleting the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
First eigenvector module, for the set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, using as described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary that generates in advance described in appearing in the eigenwert of microblogging, and whether appear at the set of described microblogging text primitive character lexical item for the lexical item of the feature lexical item dictionary that generates in advance described in judging, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0,
Sort module, for use the disaggregated model that obtains in advance by user's microblogging data automatic classification to the classification of dividing in advance;
Recommending module is foundation for take the result of automatic classification, to the user's recommended advertisements that reads microblogging data.
In the embodiment of the present invention, the information comprising than user tag due to the microblogging data of user issue has more real-time, interest preference that more can representative of consumer, and the judged result that therefore the microblogging data by analysis user obtain is more accurate, thereby the advertisement of recommending is also more accurate, and effect is also better.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of a kind of advertisement recommend method based on microblogging of providing of first embodiment of the invention;
Fig. 2 is a kind of advertisement commending system structural drawing based on microblogging that second embodiment of the invention provides;
Fig. 3 is the advertisement commending system structural drawing of the another kind that provides of second embodiment of the invention based on microblogging.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
The embodiment of the present invention, by the microblogging data of user's issue are excavated, classified, is judged this user's interest preference, and then is recommended corresponding advertisement to this user.
It is a kind of that the embodiment of the present invention provides: advertisement recommend method and system based on microblogging.
Described method comprises: the microblogging data that read user;
The microblogging data that initialization is read, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read;
Delete the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
The set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and the word frequency-reverse file frequency tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, using as described in appear at described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary of generation in advance in the eigenwert of microblogging;
Whether the lexical item of the feature lexical item dictionary generating in advance described in judgement appears in the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0;
Use the disaggregated model obtain in advance by user's microblogging data automatic classification in the classification of dividing in advance;
The result of automatic classification of take is foundation, to the user's recommended advertisements that reads microblogging data.
Described system comprises: the first data are read in module, for reading user's microblogging data;
The first data initialization module, the microblogging data that read for initialization, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read;
First Characteristic extraction module, for deleting the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
First eigenvector module, for the set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, using as described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary that generates in advance described in appearing in the eigenwert of microblogging, and whether appear at the set of described microblogging text primitive character lexical item for the lexical item of the feature lexical item dictionary that generates in advance described in judging, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0,
Sort module, for use the disaggregated model that obtains in advance by user's microblogging data automatic classification to the classification of dividing in advance;
Recommending module is foundation for take the result of automatic classification, to the user's recommended advertisements that reads microblogging data.
In the embodiment of the present invention, the information comprising than user tag due to the microblogging data of user issue has more real-time, interest preference that more can representative of consumer, and the judged result that therefore the microblogging data by analysis user obtain is more accurate, thereby the advertisement of recommending is also more accurate, and effect is also better.
For technical solutions according to the invention are described, below by specific embodiment, describe.
embodiment mono-:
Fig. 1 shows a kind of advertisement recommend method based on microblogging that first embodiment of the invention provides, and details are as follows:
Step S11, reads user's microblogging data.
In this step, can obtain in advance user's microblogging data, the microblogging data of obtaining are stored in database, need to be to certain user's microblogging data analysis time, then read this user's microblogging data.
Step S12, the microblogging data that initialization is read, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read.
In this step, every microblogging data are carried out to initialization process, such as removing the special symbols such as punctuation mark, removing non-Chinese character, participle etc., after initialization process, obtain the set of a microblogging text lexical item.
Step S13, deletes the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item.
Step S14, the set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and word frequency-reverse file frequency (term frequency-inverse document frequency of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, tf-idf) value, using as described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary that generates in advance described in appearing in the eigenwert of microblogging.
In this step, the microblogging text primitive character lexical item set of every microblogging is shone upon to feature lexical item dictionary, if the tf-idf value that the lexical item of microblogging text primitive character lexical item set at feature lexical item dictionary, is calculated this lexical item is so the eigenwert in this microblogging as this lexical item.
Step S15, whether the lexical item of the feature lexical item dictionary generating in advance described in judgement appears in the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0.
In this step, the lexical item of microblogging text primitive character lexical item set is not at feature lexical item dictionary, and this lexical item is left in the basket, if the lexical item in feature lexical item dictionary does not appear in the set of microblogging text primitive character lexical item, the eigenwert of this lexical item is 0; The microblogging text of final every microblogging is transformed into the proper vector that a dimension is 5000.
Step S16, use the disaggregated model obtain in advance by user's microblogging data automatic classification in the classification of dividing in advance.
In this step, can divide in advance according to the actual requirements plurality of classes, such as, divide in advance 12 kinds, have respectively sport category, healthy class, educational, GT grand touring, scientific and technological class, automotive-type, game class, beauty treatment, hairdressing and body shaping class, cuisines class, clothes footwear boots bag class, entertainment class, other.
Wherein, sport category comprises the contents such as competitive sports, physical culture newpapers and periodicals, sports star;
Wherein, healthy class comprises the contents such as healthy general knowledge, medicine, physical condition;
Wherein, the educational training organizations such as ,Xin navigation channel, New Orient that comprise, individual's study condition, learning intent, the content such as go abroad to study;
Wherein, GT grand touring comprises the contents such as sight spot, recreation ground, travel abroad, free walker, hotel;
Wherein, scientific and technological class comprises the contents such as mobile phone, computer, digital product;
Wherein, automotive-type comprises the contents such as automobile, automobile journal;
Wherein, game class comprises the contents such as mobile phone games, web game, online game;
Wherein, beauty treatment, hairdressing and body shaping class comprises the contents such as skin care item, cosmetics, manicure, slim, washing product;
Wherein, cuisines class comprises the contents such as food, good-for-nothing, recipe;
Wherein, entertainment class comprises the contents such as amusement circles, concert, modern drama, exhibition;
Wherein, other comprise the contents such as ownness, individual emotion, social view, life view.
Step S17, the result of automatic classification of take is foundation, to the user's recommended advertisements that reads microblogging data.
In this step, if the result of automatic classification is that user's microblogging data are included into certain class, to user, recommend the advertisement corresponding with this classification.The advertisement here comprises news, music, film, microblogging etc.
In the embodiment of the present invention, by the microblogging data of user's issue are excavated, classified, judge this user's interest preference, and then recommend corresponding advertisement to this user.The information that the microblogging data of issuing due to user comprise than user tag has more real-time, interest preference that more can representative of consumer, therefore the judged result that the microblogging data by analysis user obtain is more accurate, thereby the advertisement of recommending is also more accurate, and effect is also better.
As one embodiment of the present invention, at step S16, use the disaggregated model obtaining in advance that user's microblogging data automatic classification is comprised the steps: to the step in the classification of dividing in advance before
Steps A, read training microblogging data.
In this step, read a plurality of users' microblogging data as the microblogging data of training, to improve the accuracy of follow-up excavation as far as possible.
Step B, the classification by the described training microblogging data handmarking who reads for dividing in advance.
In this step, several makers for the class in the classification of dividing in advance, when the classification of every microblogging data of mark, use the principle that the minority is subordinate to the majority by every the microblogging data markers reading.
The training microblogging data that step C, initialization are read, to obtain the set of microblogging text lexical item, the training microblogging data that described initialization is read comprise in the special symbol of removing in the training microblogging data that read, non-Chinese character, participle.
Step D, delete the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item.
Step e, generating feature lexical item dictionary.
In this step, the step of generating feature lexical item dictionary specifically comprises: the mutual information value of calculating each lexical item in the set of microblogging text primitive character lexical item; Choose mutual information value rank in N the lexical item of front N the lexical item as feature lexical item dictionary, described N is integer, N is greater than 0.For example select 5000 lexical items that mutual information value is the highest as the lexical item of feature lexical item dictionary, the feature lexical item dictionary of generation can be arranged according to the height of mutual information value.
Step F, the lexical item set of described microblogging text primitive character and described feature lexical item dictionary are shone upon, judge whether the lexical item in the set of described microblogging text primitive character lexical item appears in described feature lexical item dictionary, and calculate the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in present described feature lexical item dictionary, using as described in appear at lexical item in the described microblogging text primitive character lexical item set in described feature lexical item dictionary in the eigenwert of microblogging.
Step G, judge whether the lexical item of described feature lexical item dictionary appears in the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary in the set of described microblogging text primitive character lexical item is labeled as to 0.
Step H, the proper vector that adopts all eigenwerts that default Algorithm for Training calculates to form, to obtain disaggregated model.
In this step, train all microblogging data characteristic of correspondence vector matrixs, during follow-up certain user's of excavation microblogging data, can directly use the result after training.
Wherein, default algorithm comprises following any algorithm: support vector machines, Naive Bayes Classification Algorithm, neural network, K close on sorting algorithm, genetic algorithm.
In the present embodiment, by analyzing the microblogging data of a large number of users, generating feature lexical item dictionary, this feature lexical item dictionary provides a normative reference for excavate certain user's interest preference later.
As one embodiment of the present invention, step S17, the result of automatic classification of take is foundation, specifically comprises: the shared number percent of every class microblogging in the microblogging of counting user to the step that reads user's recommended advertisements of microblogging data; The shared number percent of every class microblogging of statistics is mated with the label of user in microblogging data, and the shared number percent of the classification that the match is successful is double; To the user who reads microblogging data, recommend rank in the advertisement of M the classification of front M, described M is integer, and M is greater than 0.
In the present embodiment, to user's historical microblogging carry out classification count every class microblogging percentage and and this subscriber data in label mate, if label contains certain class, such microblogging percentage is double so, finally select M the classification that number percent is the highest, for example, select three classifications and recommend classification as this user's advertisement.Preferably, after a period of time, can recalculate the advertisement recommendation classification that show that this user is up-to-date.
embodiment bis-:
Fig. 2 shows the structure of a kind of advertisement commending system based on microblogging that second embodiment of the invention provides, and for convenience of explanation, only shows the part relevant to the embodiment of the present invention.
The advertisement commending system being somebody's turn to do based on microblogging can be for passing through the various information processing terminals of wired or wireless network connection server, mobile phone for example, pocket computing machine (Pocket Personal Computer, PPC), palm PC, computing machine, notebook computer, personal digital assistant (Personal Digital Assistant, PDA) etc., it can be the software unit running in these information processing terminals, the unit that hardware cell or software and hardware combine, also can be used as independently suspension member is integrated in these information processing terminals or runs in the application system of these information processing terminals, wherein:
The first data are read in module 201, for reading user's microblogging data.
The first data initialization module 202, the microblogging data that read for initialization, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise in the special symbol of removing in the microblogging data that read, non-Chinese character, participle.
First Characteristic extraction module 203, for deleting the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item.
Wherein, through the calculating of first eigenvector module 204, the microblogging data of every microblogging change the proper vector that a latitude is 5000 into the most at last.
Wherein, the classification of dividing in advance can be 12 classes, specifically, as shown in step S16, repeats no more here.
Recommending module 206 is foundation for take the result of automatic classification, to the user's recommended advertisements that reads microblogging data.
Wherein, the advertisement here comprises the contents such as news, music, film, microblogging.
In the embodiment of the present invention, by the microblogging data to reading, excavate, classification under dividing, and recommend the advertisement relevant to dividing classification to user.Because microblogging data can reflect user's interest preference in time, the judged result that therefore the microblogging data by analysis user obtain is more accurate, thereby the advertisement of recommending is also more accurate, and effect is also better.
Fig. 3 shows another structure of the advertisement commending system based on microblogging, and as another preferred embodiment of the present invention, the described advertisement commending system based on microblogging also comprises:
The second data are read in module 301, for reading training microblogging data.
Wherein, the microblogging data that the microblogging data that read are a plurality of users.
Manual sort's module 302, for the classification for dividing in advance by the described training microblogging data handmarking who reads.
The second data initialization module 303, the training microblogging data that read for initialization, to obtain the set of microblogging text lexical item, the training microblogging data that described initialization is read comprise in the special symbol of removing in the training microblogging data that read, non-Chinese character, participle.
Second Characteristic extraction module 304, for deleting the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item.
Feature lexical item dictionary generation module 305, for generating feature lexical item dictionary.
Wherein, feature lexical item dictionary generation module 305 comprises:
Mutual information value computing module, for calculating the mutual information value of each lexical item of microblogging text primitive character lexical item set.
Feature lexical item dictionary lexical item is selected module, for choose mutual information value rank in N the lexical item of front N the lexical item as feature lexical item dictionary, described N is integer, N is greater than 0.
Second Characteristic vectorization module 306, for the lexical item set of described microblogging text primitive character and described feature lexical item dictionary are shone upon, judge whether the lexical item in the set of described microblogging text primitive character lexical item appears in described feature lexical item dictionary, and calculate the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in present described feature lexical item dictionary, using as described in appear at lexical item in the described microblogging text primitive character lexical item set in described feature lexical item dictionary in the eigenwert of microblogging.And for judging whether the lexical item of described feature lexical item dictionary appears at the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary in the set of described microblogging text primitive character lexical item is labeled as to 0.
Wherein, default algorithm comprises following any algorithm:
Support vector machines, Naive Bayes Classification Algorithm, neural network, K close on sorting algorithm, genetic algorithm.
In the present embodiment, by analyzing the microblogging data of a large number of users, generating feature lexical item dictionary, this feature lexical item dictionary provides a normative reference for excavate certain user's interest preference later.
As one embodiment of the present invention, described recommending module 206 comprises:
Data statistics module, for the shared number percent of the every class microblogging of microblogging of counting user.
Data Matching module, for the shared number percent of every class microblogging of statistics is mated with the label of user in microblogging data, and the shared number percent of the classification that the match is successful is double.
Advertisement recommending module, recommends rank in the advertisement of M the classification of front M for the user to reading microblogging data, and described M is integer, and M is greater than 0.
In the present embodiment, only choose rank and recommend client in the advertisement of a front M classification, make advertisement putting more accurate not increasing client and browse on the basis of pressure.
In embodiments of the present invention, by the microblogging data of user's issue are excavated, classified, and at the label information of microblogging, judge this user's interest preference in conjunction with user, and then recommend corresponding advertisement to this user.The information that the microblogging data of issuing due to user comprise than user tag has more real-time, interest preference that more can representative of consumer, therefore the judged result that microblogging data and the label information by analysis user obtains is than only analyzing tags information is more accurate, thereby the advertisement of recommending is also more accurate, and effect is also better.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (10)
1. the advertisement recommend method based on microblogging, is characterized in that, described method comprises the steps:
Read user's microblogging data;
The microblogging data that initialization is read, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read;
Delete the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
The set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and the word frequency-reverse file frequency tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, using as described in appear at described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary of generation in advance in the eigenwert of microblogging;
Whether the lexical item of the feature lexical item dictionary generating in advance described in judgement appears in the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0;
Use the disaggregated model obtain in advance by user's microblogging data automatic classification in the classification of dividing in advance;
The result of automatic classification of take is foundation, to the user's recommended advertisements that reads microblogging data.
2. the method for claim 1, is characterized in that, the disaggregated model obtaining in advance in described use comprised the steps: user's microblogging data automatic classification before in the classification of dividing in advance
Read training microblogging;
Classification by the described training microblogging data handmarking who reads for dividing in advance;
The training microblogging data that initialization is read, to obtain the set of microblogging text lexical item, the training microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the training microblogging data that read;
Delete the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
Generating feature lexical item dictionary;
The lexical item set of described microblogging text primitive character and described feature lexical item dictionary are shone upon, judge whether the lexical item in the set of described microblogging text primitive character lexical item appears in described feature lexical item dictionary, and calculate the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in present described feature lexical item dictionary, using as described in appear at lexical item in the described microblogging text primitive character lexical item set in described feature lexical item dictionary in the eigenwert of microblogging;
Whether the lexical item that judges described feature lexical item dictionary appears in the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary in the set of described microblogging text primitive character lexical item is labeled as to 0;
The proper vector that all eigenwerts that adopt default Algorithm for Training to calculate form, to obtain disaggregated model.
3. method as claimed in claim 2, is characterized in that, the step of described generating feature lexical item dictionary specifically comprises:
Calculate the mutual information value of each lexical item in the set of microblogging text primitive character lexical item;
Choose mutual information value rank in N the lexical item of front N the lexical item as feature lexical item dictionary, described N is integer, N is greater than 0.
4. method as claimed in claim 2, is characterized in that, described default algorithm comprises following any algorithm:
Support vector machines, Naive Bayes Classification Algorithm, neural network, K close on sorting algorithm, genetic algorithm.
5. the method for claim 1, is characterized in that, described result of take automatic classification is foundation, to the step that reads user's recommended advertisements of microblogging data, specifically comprises:
The shared number percent of every class microblogging in the microblogging of counting user;
The shared number percent of every class microblogging of statistics is mated with the label of user in microblogging data, and the shared number percent of the classification that the match is successful is double;
To the user who reads microblogging data, recommend rank in the advertisement of M the classification of front M, described M is integer, and M is greater than 0.
6. the advertisement commending system based on microblogging, is characterized in that, described system comprises:
The first data are read in module, for reading user's microblogging data;
The first data initialization module, the microblogging data that read for initialization, to obtain the set of microblogging text lexical item, the microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the microblogging data that read;
First Characteristic extraction module, for deleting the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
First eigenvector module, for the set of described microblogging text primitive character lexical item and the feature lexical item dictionary generating are in advance shone upon, judge in the feature lexical item dictionary whether lexical item in the set of described microblogging text primitive character lexical item generate described in appearing in advance, and the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in the feature lexical item dictionary generating in advance described in calculating now, using as described in lexical item in described microblogging text primitive character lexical item set in the feature lexical item dictionary that generates in advance described in appearing in the eigenwert of microblogging, and whether appear at the set of described microblogging text primitive character lexical item for the lexical item of the feature lexical item dictionary that generates in advance described in judging, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary generating in advance in the set of described microblogging text primitive character lexical item is labeled as to 0,
Sort module, for use the disaggregated model that obtains in advance by user's microblogging data automatic classification to the classification of dividing in advance;
Recommending module is foundation for take the result of automatic classification, to the user's recommended advertisements that reads microblogging data.
7. system as claimed in claim 6, is characterized in that, described system also comprises:
The second data are read in module, for reading training microblogging data;
Manual sort's module, for the classification for dividing in advance by the described training microblogging data handmarking who reads;
The second data initialization module, the training microblogging data that read for initialization, to obtain the set of microblogging text lexical item, the training microblogging data that described initialization is read comprise special symbol, non-Chinese character, the participle of removing in the training microblogging data that read;
Second Characteristic extraction module, for deleting the stop words of described microblogging text lexical item set, to obtain the set of microblogging text primitive character lexical item;
Feature lexical item dictionary generation module, for generating feature lexical item dictionary;
Second Characteristic vectorization module, for the lexical item set of described microblogging text primitive character and described feature lexical item dictionary are shone upon, judge whether the lexical item in the set of described microblogging text primitive character lexical item appears in described feature lexical item dictionary, and calculate the tf-idf value of the lexical item in the described microblogging text primitive character lexical item set in present described feature lexical item dictionary, using as described in appear at lexical item in the described microblogging text primitive character lexical item set in described feature lexical item dictionary in the eigenwert of microblogging; And for judging whether the lexical item of described feature lexical item dictionary appears at the set of described microblogging text primitive character lexical item, and the eigenwert that does not appear at the lexical item of the described feature lexical item dictionary in the set of described microblogging text primitive character lexical item is labeled as to 0;
Training module, for the proper vector that adopts all eigenwerts that default Algorithm for Training calculates to form, to obtain disaggregated model.
8. system as claimed in claim 7, is characterized in that, described feature lexical item dictionary generation module comprises:
Mutual information value computing module, for calculating the mutual information value of each lexical item of microblogging text primitive character lexical item set;
Feature lexical item dictionary lexical item is selected module, for choose mutual information value rank in N the lexical item of front N the lexical item as feature lexical item dictionary, described N is integer, N is greater than 0.
9. system as claimed in claim 7, is characterized in that, described default algorithm comprises following any algorithm:
Support vector machines, Naive Bayes Classification Algorithm, neural network, K close on sorting algorithm, genetic algorithm.
10. system as claimed in claim 6, is characterized in that, described recommending module comprises:
Data statistics module, for the shared number percent of the every class microblogging of microblogging of counting user;
Data Matching module, for the shared number percent of every class microblogging of statistics is mated with the label of user in microblogging data, and the shared number percent of the classification that the match is successful is double;
Advertisement recommending module, recommends rank in the advertisement of M the classification of front M for the user to reading microblogging data, and described M is integer, and M is greater than 0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310608335.7A CN103617230B (en) | 2013-11-26 | 2013-11-26 | Method and system for advertisement recommendation based microblog |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310608335.7A CN103617230B (en) | 2013-11-26 | 2013-11-26 | Method and system for advertisement recommendation based microblog |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103617230A true CN103617230A (en) | 2014-03-05 |
| CN103617230B CN103617230B (en) | 2017-02-15 |
Family
ID=50167933
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310608335.7A Active CN103617230B (en) | 2013-11-26 | 2013-11-26 | Method and system for advertisement recommendation based microblog |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103617230B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104851026A (en) * | 2015-05-21 | 2015-08-19 | 上海宾谷网络科技有限公司 | Big data based bid native advertisement reward system for positioning user in real time, and method |
| CN104915386A (en) * | 2015-05-25 | 2015-09-16 | 中国科学院自动化研究所 | Short text clustering method based on deep semantic feature learning |
| CN104992347A (en) * | 2015-06-17 | 2015-10-21 | 北京奇艺世纪科技有限公司 | Video matching advertisement method and device |
| CN105389345A (en) * | 2015-10-26 | 2016-03-09 | 天津大学 | Short message text content classification method |
| CN105975497A (en) * | 2016-04-27 | 2016-09-28 | 清华大学 | Automatic microblog topic recommendation method and device |
| CN106339402A (en) * | 2015-07-16 | 2017-01-18 | 腾讯科技（深圳）有限公司 | Method, device and system for pushing recommended contents |
| CN106886579A (en) * | 2017-01-23 | 2017-06-23 | 北京航空航天大学 | Real-time streaming textual hierarchy monitoring method and device |
| CN107086925A (en) * | 2017-03-07 | 2017-08-22 | 珠海城市职业技术学院 | A kind of internet traffic big data analysis method based on deep learning |
| CN107169799A (en) * | 2017-05-17 | 2017-09-15 | 微梦创科网络科技（中国）有限公司 | In a kind of primary information flow generation based on social networks, throws advertisement implementation method and system |
| CN107590195A (en) * | 2017-08-14 | 2018-01-16 | 百度在线网络技术（北京）有限公司 | Textual classification model training method, file classification method and its device |
| WO2018023656A1 (en) * | 2016-08-05 | 2018-02-08 | 汤隆初 | Method for adjusting advertisement push according to usage conditions of other users, and push system |
| WO2018023658A1 (en) * | 2016-08-05 | 2018-02-08 | 汤隆初 | Method for pushing advertisement according to followed public account, and push system |
| WO2018023657A1 (en) * | 2016-08-05 | 2018-02-08 | 汤隆初 | Method for adjusting wechat public account-based advertisement push technique, and push system |
| CN108399194A (en) * | 2018-01-29 | 2018-08-14 | 中国科学院信息工程研究所 | A kind of Cyberthreat information generation method and system |
| CN109145280A (en) * | 2017-06-15 | 2019-01-04 | 北京京东尚科信息技术有限公司 | The method and apparatus of information push |
| CN109214893A (en) * | 2018-08-31 | 2019-01-15 | 深圳春沐源控股有限公司 | Method of Commodity Recommendation, recommender system and computer installation |
| CN110781303A (en) * | 2019-10-28 | 2020-02-11 | 佰聆数据股份有限公司 | Short text classification method and system |
| CN111369298A (en) * | 2020-03-09 | 2020-07-03 | 成都欧魅时尚科技有限责任公司 | Method for automatically adjusting advertisement budget based on Internet hotspot event |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101634983A (en) * | 2008-07-21 | 2010-01-27 | 华为技术有限公司 | Method and device for text classification |
| US8027977B2 (en) * | 2007-06-20 | 2011-09-27 | Microsoft Corporation | Recommending content using discriminatively trained document similarity |
| CN103324708A (en) * | 2013-06-18 | 2013-09-25 | 哈尔滨工程大学 | Method of transfer learning from long text to short text |
| CN103389981A (en) * | 2012-05-08 | 2013-11-13 | 腾讯科技（深圳）有限公司 | Network label automatic identification method and system thereof |
-
2013
- 2013-11-26 CN CN201310608335.7A patent/CN103617230B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8027977B2 (en) * | 2007-06-20 | 2011-09-27 | Microsoft Corporation | Recommending content using discriminatively trained document similarity |
| CN101634983A (en) * | 2008-07-21 | 2010-01-27 | 华为技术有限公司 | Method and device for text classification |
| CN103389981A (en) * | 2012-05-08 | 2013-11-13 | 腾讯科技（深圳）有限公司 | Network label automatic identification method and system thereof |
| CN103324708A (en) * | 2013-06-18 | 2013-09-25 | 哈尔滨工程大学 | Method of transfer learning from long text to short text |
Non-Patent Citations (1)
| Title |
|---|
| 麦艺华: ""面向中文微博的社会网络分析及应用"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104851026B (en) * | 2015-05-21 | 2018-07-17 | 上海宾谷网络科技有限公司 | Position the primary advertisement reward system and method for bidding of user in real time based on big data |
| CN104851026A (en) * | 2015-05-21 | 2015-08-19 | 上海宾谷网络科技有限公司 | Big data based bid native advertisement reward system for positioning user in real time, and method |
| CN104915386A (en) * | 2015-05-25 | 2015-09-16 | 中国科学院自动化研究所 | Short text clustering method based on deep semantic feature learning |
| CN104915386B (en) * | 2015-05-25 | 2018-04-27 | 中国科学院自动化研究所 | A kind of short text clustering method based on deep semantic feature learning |
| CN104992347A (en) * | 2015-06-17 | 2015-10-21 | 北京奇艺世纪科技有限公司 | Video matching advertisement method and device |
| CN104992347B (en) * | 2015-06-17 | 2018-12-14 | 北京奇艺世纪科技有限公司 | A kind of method and device of video matching advertisement |
| CN106339402A (en) * | 2015-07-16 | 2017-01-18 | 腾讯科技（深圳）有限公司 | Method, device and system for pushing recommended contents |
| US10885142B2 (en) | 2015-07-16 | 2021-01-05 | Tencent Technology (Shenzhen) Company Limited | Recommended content pushing method, apparatus, terminal, server, and system |
| CN105389345A (en) * | 2015-10-26 | 2016-03-09 | 天津大学 | Short message text content classification method |
| CN105975497A (en) * | 2016-04-27 | 2016-09-28 | 清华大学 | Automatic microblog topic recommendation method and device |
| WO2018023656A1 (en) * | 2016-08-05 | 2018-02-08 | 汤隆初 | Method for adjusting advertisement push according to usage conditions of other users, and push system |
| WO2018023658A1 (en) * | 2016-08-05 | 2018-02-08 | 汤隆初 | Method for pushing advertisement according to followed public account, and push system |
| WO2018023657A1 (en) * | 2016-08-05 | 2018-02-08 | 汤隆初 | Method for adjusting wechat public account-based advertisement push technique, and push system |
| CN106886579A (en) * | 2017-01-23 | 2017-06-23 | 北京航空航天大学 | Real-time streaming textual hierarchy monitoring method and device |
| CN106886579B (en) * | 2017-01-23 | 2020-01-14 | 北京航空航天大学 | Real-time streaming text grading monitoring method and device |
| CN107086925B (en) * | 2017-03-07 | 2020-04-07 | 珠海城市职业技术学院 | Deep learning-based internet traffic big data analysis method |
| CN107086925A (en) * | 2017-03-07 | 2017-08-22 | 珠海城市职业技术学院 | A kind of internet traffic big data analysis method based on deep learning |
| CN107169799A (en) * | 2017-05-17 | 2017-09-15 | 微梦创科网络科技（中国）有限公司 | In a kind of primary information flow generation based on social networks, throws advertisement implementation method and system |
| CN107169799B (en) * | 2017-05-17 | 2020-10-27 | 微梦创科网络科技（中国）有限公司 | Method and system for realizing advertisement delivery instead of native information stream based on social relationship |
| CN109145280A (en) * | 2017-06-15 | 2019-01-04 | 北京京东尚科信息技术有限公司 | The method and apparatus of information push |
| CN107590195A (en) * | 2017-08-14 | 2018-01-16 | 百度在线网络技术（北京）有限公司 | Textual classification model training method, file classification method and its device |
| CN108399194A (en) * | 2018-01-29 | 2018-08-14 | 中国科学院信息工程研究所 | A kind of Cyberthreat information generation method and system |
| CN109214893A (en) * | 2018-08-31 | 2019-01-15 | 深圳春沐源控股有限公司 | Method of Commodity Recommendation, recommender system and computer installation |
| CN110781303A (en) * | 2019-10-28 | 2020-02-11 | 佰聆数据股份有限公司 | Short text classification method and system |
| CN111369298A (en) * | 2020-03-09 | 2020-07-03 | 成都欧魅时尚科技有限责任公司 | Method for automatically adjusting advertisement budget based on Internet hotspot event |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103617230B (en) | 2017-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103617230B (en) | Method and system for advertisement recommendation based microblog | |
| US11734725B2 (en) | Information sending method, apparatus and system, and computer-readable storage medium | |
| CN103678335B (en) | The method of method, apparatus and the commodity navigation of commodity sign label | |
| CN111784455A (en) | Article recommendation method and recommendation equipment | |
| US20160132601A1 (en) | Hybrid Explanations In Collaborative Filter Based Recommendation System | |
| CN106503015A (en) | A kind of method for building user's portrait | |
| CN106294500B (en) | Content item pushing method, device and system | |
| CN106105096A (en) | System and method for continuous social communication | |
| CN102576438A (en) | Method and apparatus for executing a recommendation | |
| CN103365867A (en) | Method and device for emotion analysis of user evaluation | |
| Zhang et al. | Multimodal marketing intent analysis for effective targeted advertising | |
| CN111898031A (en) | Method and device for obtaining user portrait | |
| CN103207914A (en) | Preference vector generation method and preference vector generation system based on user feedback evaluation | |
| JP5754854B2 (en) | Contributor analysis apparatus, program and method for analyzing poster profile information | |
| CN104050243B (en) | It is a kind of to search for the network search method combined with social activity and its system | |
| CN106776859A (en) | Mobile solution App commending systems based on user preference | |
| CN115244547A (en) | Automatically and intelligently exploring design spaces | |
| CN113934941A (en) | User recommendation system and method based on multi-dimensional information | |
| CN108984711A (en) | A kind of personalized APP recommended method based on layering insertion | |
| KR20130053448A (en) | Search device, search method, search program, and computer-readable memory medium for recording search program | |
| CN116823410B (en) | Data processing method, object processing method, recommending method and computing device | |
| CN106383857A (en) | Information processing method and electronic equipment | |
| US20230351473A1 (en) | Apparatus and method for providing user's interior style analysis model on basis of sns text | |
| Singgalen | Implementing Rapid Application Development (RAD) for Statistical Analysis of Tourism and Travel Vlog Content | |
| CN111026957B (en) | Recommendation system and method based on multidimensional similarity |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20140305 Assignee: Dongguan Shengnuolin Sports Products Co.,Ltd. Assignor: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Contract record no.: X2023980037877 Denomination of invention: A Weibo based advertising recommendation method and system Granted publication date: 20170215 License type: Common License Record date: 20230712 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20140305 Assignee: Shenzhen Huayun Xingchuang Cultural Technology Co.,Ltd. Assignor: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Contract record no.: X2023980043804 Denomination of invention: A Weibo based advertising recommendation method and system Granted publication date: 20170215 License type: Common License Record date: 20231018 Application publication date: 20140305 Assignee: Shenzhen Xingfei Software Technology Co.,Ltd. Assignor: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Contract record no.: X2023980043566 Denomination of invention: A Weibo based advertising recommendation method and system Granted publication date: 20170215 License type: Common License Record date: 20231016 |
|
| EE01 | Entry into force of recordation of patent licensing contract |