CN108664585A - Word method is selected in a kind of advertisement based on big data - Google Patents
Word method is selected in a kind of advertisement based on big data Download PDFInfo
- Publication number
- CN108664585A CN108664585A CN201810425266.9A CN201810425266A CN108664585A CN 108664585 A CN108664585 A CN 108664585A CN 201810425266 A CN201810425266 A CN 201810425266A CN 108664585 A CN108664585 A CN 108664585A
- Authority
- CN
- China
- Prior art keywords
- word
- phrase
- app
- weight
- big data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical fields that word is selected in advertisement, and in particular to word method is selected in a kind of advertisement based on big data;The technical issues of solution is:It provides and a kind of the higher advertisement based on big data of word efficiency is selected to select word method;The technical solution used for:Word method, including auto recommending method are selected in a kind of advertisement based on big data, the auto recommending method, including:Establish the relevant selected phrase databases of APP;Extract all candidate keywords;The weight of each candidate keywords is calculated, and is ranked up by weight;Word quantity N is selected in reception;It is pushed out N number of dispensing keyword;Further include:Manual adding method, the manual adding method, including:According to the affiliated industries of APP are launched, minute mark label are shown candidate word;Receive screening instruction;It is pushed out multiple dispensing words.
Description
Technical field
The invention belongs to the technical fields that word is selected in advertisement, and in particular to word method is selected in a kind of advertisement based on big data.
Background technology
Apple search advertisements are the APPle store application modes risen in 2017, and core is exactly to select most
The keyword of potentiality more shows and downloads to be brought with limited budget.
The primary backstage of apple provides two kinds and selects word mode:(1) addition manually;Single input uploads table, (2) apple one
Key is recommended;All there is significant limitation in both modes.
1. the limitation of addition keyword manually:
Addition keyword problems faced is how selection keyword manually, and general optimization teacher can be by third party's data
Statistical tool downloads initial data from third-party platform, then by tools such as excel, after a series of screenings, selects symbol
Close the keyword of dispensing demand;The keyword addition page for launching backstage is then switched to, the keyword selected is uploaded;Entirely
Process is not only complicated, simultaneously because the limitation of third party's data, keeps the dimension for obtaining word very single, it is difficult to select comprehensively, and with
The keyword that advertiser's application more suits, to directly affect the Data Representation of dispensing.
2. one key of apple is recommended to select the limitation of word:
The keyword of recommendation not only limited amount, while how Classification Management and operation are carried out also to keyword and become such
Select the pain spot of word mode;General different types of keyword, initial bid has certain difference, but one key of apple is selected to push away
Such demand can not be realized when recommending.
Invention content
The present invention overcomes the shortcomings of the prior art, technical problem to be solved to be:There is provided it is a kind of select word efficiency compared with
Word method is selected in the high advertisement based on big data.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention is:
Word method, including auto recommending method are selected in a kind of advertisement based on big data, the auto recommending method, including:
Establish the relevant selected phrase databases of APP;Extract all candidate keywords;The weight of each candidate keywords is calculated, and by power
It is ranked up again;Word quantity N is selected in reception;It is pushed out N number of dispensing keyword.
Preferably, further include:Manual adding method, the manual adding method, including:According to launch the affiliated industries of APP,
Minute mark label are shown candidate word;Receive screening instruction;It is pushed out multiple dispensing words.
Preferably, further include:Translation on line is carried out to launching word.
Preferably, described to establish the relevant selected phrase databases of APP, it specifically includes:
S1011 extracts phrase important in title, description according to the title of APP, description;
IDF dictionaries are calculated according to the phrase extracted in S1012;
S1013, in conjunction with IDF dictionaries, calculates the TF-IDF values of each phrase, obtains after removing the stop-word in title, description
To the selected phrase of Weight;
S104 extends to the major category of APP in selected phrase;Wherein:The weight of the major category of APP is using expansion position
The weight of periphery word;
S105 is ranked up selected phrase according to weight, forms the relevant selected phrase databases of APP.
Preferably, described to extract in all candidate keywords, the candidate keywords, including:In selected phrase database
All phrases, and/or the phrase launched in history, and/or the phrase that covered.
Preferably, described to calculate the weight of each candidate keywords, and be ranked up by weight, it specifically includes:
The phrase for specifying number is chosen in selected phrase database;
Term vector model is trained by word2vec algorithms;
According to trained term vector model, the term vector of selected several phrases is calculated separately out;
Traverse all candidate keywords, using trained term vector model calculate the words of each candidate keywords to
Amount;
The cosine similarity between candidate keywords term vector and selected phrase vector is calculated separately, and seeks weighted average
Value;
According to the search temperature of candidate keywords and the weighted average of calculating, the power of each candidate keywords is calculated
Weight;
Candidate keywords are ranked up from big to small by weight.
Preferably, described to be chosen in selected phrase database in the phrase for specifying number, the specified number is 7.
Preferably, described according to the affiliated industries of APP are launched, minute mark label are shown candidate word, specifically include:
The keyword and competing product for calculating the hot word, covering itself of launching the affiliated industries of APP cover word, form candidate word;
According to searchable index, popularity, searching position, number of applications of bidding, candidate word minute mark label are shown;
The reception screening instruction, including:
Receive one or more screening conditions input instructions;Wherein:
The screening conditions include:Keyword title, application site, is bidded using number at search temperature in the recent period.
The present invention has the advantages that compared with prior art:
1, the present invention in auto recommending method, can choose with the relevant phrase databases of APP, pass through extract candidate key
Word, weight calculation capture out the dispensing keyword of specified quantity, and realize automation selects word, selects word efficiency higher.
2, the manual adding method in the present invention, by the excavation of keyword, analysis, screening, addition, one-stop operation,
It improves and selects word efficiency.
3, the manual adding method in the present invention, can be according to the affiliated industries of APP be launched, and minute mark label open up candidate word
To show, the hot word of the affiliated industries of APP is launched in calculating, the keyword and competing product of covering in itself cover word, formation candidate word, and according to
Searchable index, popularity, searching position, number of applications of bidding are shown candidate word minute mark label, and candidate keywords classification is bright
Really, reference value is high.
4, the present invention provides the keyword database of various dimensions, data by the crawl to mass data progress keyword
It captures efficient.
5, in the present invention, by carrying out translation on line to launching word, solving rare foreign languages keyword and launching obstacle, use
Translation on line function aid decision, is rapidly completed and selects word.
Description of the drawings
The present invention will be further described in detail below in conjunction with the accompanying drawings;
Fig. 1 is the flow diagram that word method is selected in a kind of advertisement based on big data that the embodiment of the present invention one provides;
Fig. 2 is the flow diagram that word method is selected in a kind of advertisement based on big data provided by Embodiment 2 of the present invention.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments;Based on the embodiments of the present invention, ordinary skill people
The every other embodiment that member is obtained without creative efforts, shall fall within the protection scope of the present invention.
Fig. 1 is the flow diagram that word method is selected in a kind of advertisement based on big data that the embodiment of the present invention one provides, such as
Shown in Fig. 1, word method, including auto recommending method are selected in a kind of advertisement based on big data, the auto recommending method, including:
Establish the relevant selected phrase databases of APP;Extract all candidate keywords;The weight of each candidate keywords is calculated, and by power
It is ranked up again;Word quantity N is selected in reception;It is pushed out N number of dispensing keyword.
Specifically, described to establish the relevant selected phrase databases of APP, including:
S1011 extracts phrase important in title, description according to the title of APP, description;
IDF dictionaries are calculated according to the phrase extracted in S1012;
S1013, in conjunction with IDF dictionaries, calculates the TF-IDF values of each phrase, obtains after removing the stop-word in title, description
To the selected phrase of Weight;
S104 extends to the major category of APP in selected phrase;Wherein:The weight of the major category of APP is using expansion position
The weight of periphery word;
S105 is ranked up selected phrase according to weight, forms the relevant selected phrase databases of APP.
Further, described to extract in all candidate keywords, the candidate keywords, including:Selected phrase database
In all phrases, and/or the phrase launched in history, and/or the phrase that covered.
Further, described to calculate the weight of each candidate keywords, and be ranked up by weight, it specifically includes:
The phrase for specifying number is chosen in selected phrase database;Term vector model is trained by word2vec algorithms;According to trained word
Vector model calculates separately out the term vector of selected several phrases;Traverse all candidate keywords, using trained word to
Amount model calculates the term vector of each candidate keywords;It calculates separately between candidate keywords term vector and selected phrase vector
Cosine similarity, and seek weighted average;According to the search temperature of candidate keywords and the weighted average of calculating, calculate
Go out the weight of each candidate keywords;Candidate keywords are ranked up from big to small by weight.
Described to be chosen in selected phrase database in the phrase for specifying number, the specified number is 7.
Auto recommending method in the present invention, can choose with the relevant phrase databases of APP, pass through extract candidate key
Word, weight calculation capture out the dispensing keyword of specified quantity, and realize automation selects word, have expanded the magnitude for selecting word, choosing
Word efficiency is higher.
Fig. 2 is the flow diagram that word method is selected in a kind of advertisement based on big data that the embodiment of the present invention one provides, such as
Shown in Fig. 2, on the basis of embodiment one, further include:Manual adding method, the manual adding method, including:According to dispensing
The affiliated industries of APP, minute mark label are shown candidate word;Receive screening instruction;It is pushed out multiple dispensing words;It is carried out to launching word
Translation on line.
Specifically, described according to the affiliated industries of APP are launched, minute mark label are shown candidate word, specifically include:It calculates and throws
The hot word of the affiliated industries of APP, the keyword of covering itself and competing product covering word are put, candidate word is formed;According to searchable index, stream
Row degree, searching position, number of applications of bidding are shown candidate word minute mark label;The reception screening instruction, including:It receives
One or more screening conditions input instructions;Wherein:The screening conditions include:Keyword title searches for temperature, using position
It sets, bidded in the recent period using number.
Manual adding method in the present invention, by the excavation of keyword, analysis, screening, addition, one-stop operation carries
It is high to select word efficiency;Can be according to the affiliated industries of APP to be launched, minute mark label are shown candidate word, calculate and launch the affiliated rows of APP
The hot word of industry, the keyword of covering itself and competing product cover word, form candidate word, and according to searchable index, popularity, search
Position, number of applications of bidding are shown candidate word minute mark label, and candidate keywords classification is clear, and reference value is high;By right
Mass data carries out the crawl of keyword, provides the keyword database of various dimensions, data grabber is efficient.
In addition, in the present invention, by carrying out translation on line to launching word, solving rare foreign languages keyword and launching obstacle,
With online interpretative function aid decision, it is rapidly completed and selects word.
In the present embodiment, IDF dictionaries are:Inverse document frequency dictionary, stop-word are:Stopwords, i.e.,:Similar a, an,
The word of the not no practical significance such as the.
Finally it should be noted that:The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations;To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that:Its according to
So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into
Row equivalent replacement;And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (8)
1. word method is selected in a kind of advertisement based on big data, it is characterised in that:Including auto recommending method, the automatic recommendation side
Method, including:
Establish the relevant selected phrase databases of APP;
Extract all candidate keywords;
The weight of each candidate keywords is calculated, and is ranked up by weight;
Word quantity N is selected in reception;
It is pushed out N number of dispensing keyword.
2. word method is selected in a kind of advertisement based on big data according to claim 1, it is characterised in that:Further include:Manually
Adding method, the manual adding method, including:
According to the affiliated industries of APP are launched, minute mark label are shown candidate word;
Receive screening instruction;
It is pushed out multiple dispensing words.
3. word method is selected in a kind of advertisement based on big data according to claim 2, it is characterised in that:Further include:
Translation on line is carried out to launching word.
4. word method is selected in a kind of advertisement based on big data according to claim 1, it is characterised in that:
It is described to establish the relevant selected phrase databases of APP, it specifically includes:
S1011 extracts phrase important in title, description according to the title of APP, description;
IDF dictionaries are calculated according to the phrase extracted in S1012;
S1013, in conjunction with IDF dictionaries, calculates the TF-IDF values of each phrase, obtains band after removing the stop-word in title, description
The selected phrase of weight;
S104 extends to the major category of APP in selected phrase;Wherein:The weight of the major category of APP is using expansion position periphery
The weight of word;
S105 is ranked up selected phrase according to weight, forms the relevant selected phrase databases of APP.
5. word method is selected in a kind of advertisement based on big data according to claim 1, it is characterised in that:It is described to extract institute
Have in candidate keywords, the candidate keywords, including:
All phrases in selected phrase database, and/or the phrase launched in history, and/or the phrase that covered.
6. word method is selected in a kind of advertisement based on big data according to claim 1, it is characterised in that:It is described to calculate each
The weight of candidate keywords, and be ranked up by weight, it specifically includes:
The phrase for specifying number is chosen in selected phrase database;
Term vector model is trained by word2vec algorithms;
According to trained term vector model, the term vector of selected several phrases is calculated separately out;
All candidate keywords are traversed, the term vector of each candidate keywords is calculated using trained term vector model;
The cosine similarity between candidate keywords term vector and selected phrase vector is calculated separately, and seeks weighted average;
According to the search temperature of candidate keywords and the weighted average of calculating, the weight of each candidate keywords is calculated;
Candidate keywords are ranked up from big to small by weight.
7. word method is selected in a kind of advertisement based on big data according to claim 6, it is characterised in that:It is described in selected word
It is chosen in group library in the phrase for specifying number, the specified number is 7.
8. word method is selected in a kind of advertisement based on big data according to claim 2, it is characterised in that:
It is described according to the affiliated industries of APP are launched, minute mark label are shown candidate word, specifically include:
The keyword and competing product for calculating the hot word, covering itself of launching the affiliated industries of APP cover word, form candidate word;
According to searchable index, popularity, searching position, number of applications of bidding, candidate word minute mark label are shown;
The reception screening instruction, including:
Receive one or more screening conditions input instructions;Wherein:
The screening conditions include:Keyword title, application site, is bidded using number at search temperature in the recent period.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810425266.9A CN108664585A (en) | 2018-05-07 | 2018-05-07 | Word method is selected in a kind of advertisement based on big data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810425266.9A CN108664585A (en) | 2018-05-07 | 2018-05-07 | Word method is selected in a kind of advertisement based on big data |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108664585A true CN108664585A (en) | 2018-10-16 |
Family
ID=63778352
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810425266.9A Pending CN108664585A (en) | 2018-05-07 | 2018-05-07 | Word method is selected in a kind of advertisement based on big data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108664585A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109947913A (en) * | 2019-01-26 | 2019-06-28 | 浙江乾冠信息安全研究院有限公司 | A kind of grey black produces the keyword lookup method of popularization |
CN110516030A (en) * | 2019-08-26 | 2019-11-29 | 北京百度网讯科技有限公司 | It is intended to determination method, apparatus, equipment and the computer readable storage medium of word |
CN110866106A (en) * | 2019-10-10 | 2020-03-06 | 重庆金融资产交易所有限责任公司 | Text recommendation method and related equipment |
CN113538062A (en) * | 2021-07-28 | 2021-10-22 | 福州果集信息科技有限公司 | Method for reversely deducing bid words purchased by commodity promotion notes |
CN113836307A (en) * | 2021-10-15 | 2021-12-24 | 国网北京市电力公司 | Power supply service work order hotspot discovery method, system and device and storage medium |
CN116645147A (en) * | 2023-07-27 | 2023-08-25 | 深圳千亚商务科技有限公司 | Automatic advertising method and system based on Amazon platform |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102043843A (en) * | 2010-12-08 | 2011-05-04 | 百度在线网络技术(北京)有限公司 | Method and obtaining device for obtaining target entry based on target application |
CN103377232A (en) * | 2012-04-25 | 2013-10-30 | 阿里巴巴集团控股有限公司 | Headline keyword recommendation method and system |
CN105893478A (en) * | 2016-03-29 | 2016-08-24 | 广州华多网络科技有限公司 | Tag extraction method and equipment |
CN106227735A (en) * | 2016-07-11 | 2016-12-14 | 苏州天梯卓越传媒有限公司 | A kind of word cloud Topic Selection for Publishing Industry and system |
CN107608983A (en) * | 2016-07-11 | 2018-01-19 | 阿里巴巴集团控股有限公司 | Heading message optimization method, device, equipment and system |
-
2018
- 2018-05-07 CN CN201810425266.9A patent/CN108664585A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102043843A (en) * | 2010-12-08 | 2011-05-04 | 百度在线网络技术(北京)有限公司 | Method and obtaining device for obtaining target entry based on target application |
CN103377232A (en) * | 2012-04-25 | 2013-10-30 | 阿里巴巴集团控股有限公司 | Headline keyword recommendation method and system |
CN105893478A (en) * | 2016-03-29 | 2016-08-24 | 广州华多网络科技有限公司 | Tag extraction method and equipment |
CN106227735A (en) * | 2016-07-11 | 2016-12-14 | 苏州天梯卓越传媒有限公司 | A kind of word cloud Topic Selection for Publishing Industry and system |
CN107608983A (en) * | 2016-07-11 | 2018-01-19 | 阿里巴巴集团控股有限公司 | Heading message optimization method, device, equipment and system |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109947913A (en) * | 2019-01-26 | 2019-06-28 | 浙江乾冠信息安全研究院有限公司 | A kind of grey black produces the keyword lookup method of popularization |
CN110516030A (en) * | 2019-08-26 | 2019-11-29 | 北京百度网讯科技有限公司 | It is intended to determination method, apparatus, equipment and the computer readable storage medium of word |
CN110866106A (en) * | 2019-10-10 | 2020-03-06 | 重庆金融资产交易所有限责任公司 | Text recommendation method and related equipment |
CN113538062A (en) * | 2021-07-28 | 2021-10-22 | 福州果集信息科技有限公司 | Method for reversely deducing bid words purchased by commodity promotion notes |
CN113538062B (en) * | 2021-07-28 | 2024-05-07 | 福州果集信息科技有限公司 | Method for reversely pushing bid words purchased by commodity popularization notes |
CN113836307A (en) * | 2021-10-15 | 2021-12-24 | 国网北京市电力公司 | Power supply service work order hotspot discovery method, system and device and storage medium |
CN113836307B (en) * | 2021-10-15 | 2024-02-20 | 国网北京市电力公司 | Power supply service work order hot spot discovery method, system, device and storage medium |
CN116645147A (en) * | 2023-07-27 | 2023-08-25 | 深圳千亚商务科技有限公司 | Automatic advertising method and system based on Amazon platform |
CN116645147B (en) * | 2023-07-27 | 2024-01-23 | 深圳千亚商务科技有限公司 | Automatic advertising method and system based on Amazon platform |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108664585A (en) | Word method is selected in a kind of advertisement based on big data | |
CN105488024B (en) | The abstracting method and device of Web page subject sentence | |
CN102253982B (en) | Query suggestion method based on query semantics and click-through data | |
CN105320772B (en) | A kind of association paper querying method of patent duplicate checking | |
CN102033949B (en) | Correction-based K nearest neighbor text classification method | |
CN101493819B (en) | Method for optimizing detection of search engine cheat | |
CN103605665A (en) | Keyword based evaluation expert intelligent search and recommendation method | |
CN106156204A (en) | The extracting method of text label and device | |
CN105653706A (en) | Multilayer quotation recommendation method based on literature content mapping knowledge domain | |
CN105975596A (en) | Query expansion method and system of search engine | |
CN103744956B (en) | A kind of diversified expanding method of key word | |
Aklouche et al. | Query Expansion Based on NLP and Word Embeddings. | |
CN105117487A (en) | Book semantic retrieval method based on content structures | |
CN108334610A (en) | A kind of newsletter archive sorting technique, device and server | |
CN105069647A (en) | Improved method for extracting evaluation object in Chinese commodity review | |
CN110555154B (en) | Theme-oriented information retrieval method | |
CN101763395A (en) | Method for automatically generating webpage by adopting artificial intelligence technology | |
CN107688621A (en) | The optimization method and system of a kind of official documents and correspondence | |
CN108491407B (en) | Code retrieval-oriented query expansion method | |
CN109408100A (en) | A kind of software defect information fusion method based on multi-source data | |
CN107133274A (en) | A kind of distributed information retrieval set option method based on figure knowledge base | |
CN116644174A (en) | Resource classification screening method and system based on technological policy | |
Yang et al. | Fault-tolerant learning for term extraction | |
Kang et al. | Automated mapping of product features mined from online customer reviews to engineering product characteristics | |
CN109189915A (en) | A kind of information retrieval method based on depth relevant matches model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181016 |
|
RJ01 | Rejection of invention patent application after publication |