CN110717104B - Keyword advertisement putting automatic negative keyword method and device - Google Patents
Keyword advertisement putting automatic negative keyword method and device Download PDFInfo
- Publication number
- CN110717104B CN110717104B CN201910962210.1A CN201910962210A CN110717104B CN 110717104 B CN110717104 B CN 110717104B CN 201910962210 A CN201910962210 A CN 201910962210A CN 110717104 B CN110717104 B CN 110717104B
- Authority
- CN
- China
- Prior art keywords
- negative
- keywords
- accurate
- keyword
- phrase
- 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.)
- Active
Links
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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
- G06Q30/0256—User search
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0277—Online advertisement
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method and a device for automatically negating keywords in keyword advertisement delivery, wherein the method comprises the following steps: automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database; determining an exact negative keyword set; extracting available common root words in the accurate negative keyword set as a phrase negative keyword set; rejecting invalid accurate negative keywords in the accurate negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set; uploading the accurate negative keyword set and the phrase negative keyword set to a corresponding account of the releasing platform. The method and the device for automatically negating the keywords in the keyword advertisement delivery can complete algorithm processing and uploading updating of the automatically negating keywords, reduce consumption of manpower and material resources and improve operation efficiency. And accurate negative keywords can be mined, so that the updated negative keywords can have the effect of filtering invalid search words, and the invalid consumption is reduced.
Description
Technical Field
The invention relates to the technical field of advertisement putting, in particular to a keyword advertisement putting automatic keyword negation method and device.
Background
Keyword advertisement marketing of a search engine refers to that after a user inputs search words on the search engine, the search engine displays keyword advertisements put by advertisers according to keyword bidding and matching rules, and charges are charged according to the modes of exposure, clicking and the like of the advertisements. However, when the advertiser puts the keyword advertisement, the user is not expected to search the keyword advertisement by some search words. If the keyword advertisement of 'credit card transaction' is released, the advertiser does not want to show the advertisement to the users who search for 'credit card is overdue', and the like, which have obvious difference from the release target. In this regard, the search engine may filter the search terms by adding negative keywords. The negative keywords of the search engine can be divided into phrase negative keywords and precise negative keywords, wherein the phrase negative keywords are used for filtering out all search terms containing the phrase, and the precise negative keywords are used for filtering out search terms completely consistent with the precise negative keywords.
At present, negative keywords are mainly added to an account in a manual mode, and specific operations include that an advertisement publisher manually checks existing search terms in a search engine advertisement publishing background; an advertisement publisher screens invalid search words from existing search words to serve as accurate negative keywords; further, the advertiser may summarize common phrases in the invalid search terms as phrase negative keywords. However, the above operations require a lot of working time and labor cost for advertisement publishers, and the manual operation efficiency is low, and the advertisement advertisers cannot add the advertisement at any time 24 hours a day.
Disclosure of Invention
In order to solve the technical problems in the prior art, the invention provides a method and a device for automatically rejecting keywords in keyword advertisement delivery. The specific technical scheme is as follows:
in a first aspect, an automated negative keyword method for keyword advertisement delivery is provided, the method comprising: automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database; determining an exact negative keyword set; extracting available common roots in the accurate negative keyword set to serve as a phrase negative keyword set; rejecting invalid accurate negative keywords in the accurate negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set; and uploading the accurate negative keyword set and the phrase negative keyword set to a corresponding account of the launching platform.
In one possible design, the method further includes: and automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time, storing the keywords, the negative keywords and the search word data in a database, and automatically adding binding if a negative keyword universal package is not added.
In one possible design, the determining the set of exact negative keywords comprises: inquiring whether a negative keyword white list exists or not, and if so, determining search terms except the negative keyword white list as an accurate negative keyword set; if no negative keyword white list exists, screening out search words with semantic relevance lower than a threshold value through a TF-IDF algorithm, adding the search words to an accurate negative keyword set, screening out search words with negative emotion through a Bayesian classification model, and adding the search words to the accurate negative keyword set.
In one possible design, the screening out search terms with semantic relevance lower than a threshold value through the TF-IDF algorithm, and adding the search terms to the precise negative keyword set, includes: taking the keyword set as a target keyword text, and performing word segmentation by using a hidden Markov model to obtain a keyword root set; performing word segmentation on the search word text by using a hidden Markov model to obtain a search word root set; calculating the word frequency TF of the root word of the search word by the following formula Iω，
Calculating the inverse file frequency IDF by the following formula IIω：
Obtaining TF-TDF value TF-IDF of search term by the following formula IIIω：
And screening out search words with TF-IDF values lower than a threshold value, and adding the search words to the accurate negative keyword set.
In one possible design, the method screens out search terms with negative emotions through a Bayesian classification model, and adds the search terms with negative emotions to an accurate negative keyword set, and comprises the following steps: loading an emotion analysis dictionary; calculating the probability that the search word belongs to the negative emotion through the following formula four:
where P is the probability that the search term belongs to a negative emotion, count (d)iAnd C) search for the word root d for negative feelingsiNumber of occurrences, VcThe total number of negative emotion root words is n, and the total number of all root words is n;
adding the search words with the negative probability higher than a threshold value to the accurate negative keyword set.
In one possible design, the extracting available common roots in the set of exact negative keywords as the set of phrase negative keywords comprises:
taking the accurate negative keyword set as a text, and performing word segmentation by using a hidden Markov model to obtain an accurate negative keyword root set;
extracting common roots in the accurate negative keyword root set, and constructing a target function:
minα·countNeg(x)-β·countexact(x)，
wherein χ is the common root combination, countNegTo share the root number, countexactCovering the number of the accurate negative keywords for the common root, wherein alpha and beta are adjustment coefficients;
and finding the optimal root combination by using a gradient descent method to serve as a phrase negative keyword set.
In one possible design, the rejecting invalid exact negative keywords in the exact negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set includes: taking the accurate negative keywords in the accurate negative keyword set, which are the same as the account keywords, as invalid accurate negative keywords to be removed; and eliminating the phrase negative keywords contained by the account keywords in the phrase negative keyword set as invalid phrase negative keywords.
In a possible design, the method further includes, after eliminating the invalid precise negative keywords in the precise negative keyword set and eliminating the invalid phrase negative keywords in the phrase negative keyword set, if the launching platform has a limit on the number of the added negative keywords, preferentially negating the high-frequency and high-click search terms.
In one possible design, the preferentially negating high frequency, high click search terms includes:
counting the total number of the accurate negative keywords as ExactNegNum, wherein the maximum addable value of the accurate negative keywords is MaxExactNegNum, the total number of the phrase negative keywords is NegNum, and the maximum addable value of the phrase negative keywords is MaxNegNum;
if the exact NegNum is larger than the MaxExactNegNum, counting the length of each accurate negative keyword of the account, calculating the average length theta and the standard deviation sigma of the length, deleting the accurate negative keywords with the length larger than (theta +3 sigma), and updating the exact NegNum;
if the exact NegNum is larger than the MaxExactNegNum, counting the click times click _ time of the search words of the account in the past n days, removing the accurate negative keywords of which the click _ time is less than the threshold value from the accurate negative keyword set, and updating the exact NegNum;
if the exact NegNum is larger than the MaxExactNegNum, counting the occurrence times impresson _ time of the search words of the past n days of the account, removing the accurate negative keywords of which the impresson _ time is less than the threshold value from the accurate negative keyword set, and updating the exact NegNum;
if the exact NegNum is more than MaxExactNegNum, the number of the accurate negative keywords to be deleted is DeleteExactNegNum which is exact NegNum-MaxExactNegNum, all the accurate negative keywords are sorted according to the length, and the DeleteExactNegNum accurate negative keywords with the longest length are deleted from the accurate negative keyword set;
if NegNum is larger than MaxNegNum, the number of the phrase negative keywords to be deleted is DeleteNegNum which is NegNum-MaxMegNum, all the phrase negative keywords are sorted according to length, and the DeleteNegNum phrase negative keywords with the longest length are deleted from the phrase negative keyword set.
In a second aspect, an apparatus for automatically negating keywords in keyword advertisement delivery is provided, the apparatus comprising: the acquisition module is used for automatically acquiring the keywords, the negative keywords and the search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database; a determination module for determining a set of exact negative keywords; the extraction module is used for extracting an available common root word in the accurate negative keyword set to serve as a phrase negative keyword set; the rejecting module is used for rejecting invalid accurate negative keywords in the accurate negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set; and the uploading module is used for uploading the accurate negative keyword set and the phrase negative keyword set to the corresponding account of the releasing platform.
The technical scheme of the invention has the following main advantages:
the method for automatically rejecting the keywords in the keyword advertisement delivery can detect and analyze the advertisement account in real time and uninterruptedly, further automatically excavate and add the rejection keywords, can complete algorithm processing and uploading updating of the automatic rejection keywords within seconds, reduces consumption of manpower and material resources, and improves operation efficiency. And accurate negative keywords can be mined, so that the updated negative keywords can have the effect of filtering invalid search words, and the invalid consumption of advertisers is reduced.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart of an automated negative keyword method for keyword advertisement delivery according to an embodiment of the present invention.
Fig. 2 is a flowchart of an automated negative keyword method for keyword advertisement delivery according to another embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the specific embodiments of the present invention and the accompanying drawings. It is to be understood that the described embodiments are merely a few embodiments of the invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
The technical scheme provided by the embodiment of the invention is described in detail below with reference to the accompanying drawings.
Referring to fig. 1, an embodiment of the present invention provides an automatic negative keyword advertisement delivery method, where the method includes:
and S101, automatically acquiring the keywords, the negative keywords and the search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database.
S102, determining an accurate negative keyword set.
S103, extracting available common roots in the accurate negative keyword set to serve as a phrase negative keyword set.
S104, eliminating invalid accurate negative keywords in the accurate negative keyword set, and eliminating invalid phrase negative keywords in the phrase negative keyword set.
And S105, uploading the accurate negative keyword set and the phrase negative keyword set to an account corresponding to the release platform.
The method for automatically rejecting the keywords in the keyword advertisement delivery provided by the embodiment of the invention can detect and analyze the advertisement account in real time without interruption so as to automatically mine and add the rejection keywords, can complete the algorithm processing and uploading updating of the automatic rejection keywords within seconds, reduces the consumption of manpower and material resources and improves the operation efficiency. And accurate negative keywords can be mined, so that the updated negative keywords can have the effect of filtering invalid search words, and the invalid consumption of advertisers is reduced.
Referring to fig. 2, another method for automatically negating keywords in a keyword advertisement according to an embodiment of the present invention further includes:
s201, automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database.
S202, if the negative keyword universal package is not added, automatically adding the binding. If not, directly entering the next step.
S203, inquiring whether a negative keyword white list exists or not, and if so, determining search terms except the negative keyword white list as an accurate negative keyword set. The process advances to step S206.
S204, if the negative keyword white list does not exist, screening out the search words with semantic relevance lower than a threshold value through a TF-IDF algorithm, and adding the search words to the accurate negative keyword set. In the step, other algorithms such as TextRank and the like can be selected for semantic relevance calculation.
Specifically, a TF-IDF algorithm is used for screening out search terms with semantic relevance lower than a threshold value, and the search terms are added to an accurate negative keyword set, and the method comprises the following steps:
and taking the keyword set as a target keyword text, and performing word segmentation by using a hidden Markov model to obtain a keyword root set.
And performing word segmentation on the search word text by using a hidden Markov model to obtain a search word root set.
Calculating the word frequency TF of the root word of the search word by the following formula Iω，
Calculating the inverse file frequency IDF by the following formula IIω：
Obtaining TF-TDF value TF-IDF of search term by the following formula IIIω：
And screening out search words with TF-IDF values lower than a threshold value, and adding the search words to the accurate negative keyword set.
S205, screening out search terms with negative emotions through a Bayesian classification model, and adding the search terms to an accurate negative keyword set. In this step, the bayesian classification model may be replaced by another classification algorithm such as the xgboost algorithm.
Specifically, screening out search terms with negative emotion through a Bayesian classification model, and adding the search terms into an accurate negative keyword set, wherein the method comprises the following steps:
and loading the emotion analysis dictionary.
Calculating the probability that the search word belongs to the negative emotion through the following formula four:
in the formula, P is the probability that the search word belongs to the negative emotion, and count (di, C) is the root word d of the negative emotion search wordiNumber of occurrences, VcThe total number of negative emotion root words is n, and the total number of all root words is n;
adding the search words with the negative probability higher than a threshold value to the accurate negative keyword set.
And S206, after the accurate negative keyword set is determined through the step S204 or the steps S205 and S206, extracting the available common root words in the accurate negative keyword set as a phrase negative keyword set.
Specifically, extracting the available common root word in the accurate negative keyword set as the phrase negative keyword set comprises the following steps:
taking the accurate negative keyword set as a text, and performing word segmentation by using a hidden Markov model to obtain an accurate negative keyword root set;
extracting common roots in the accurate negative keyword root set, and constructing a target function:
min α·countNeg(x)-β·countexact(x)，
wherein x is a common root combination, countNegTo share the root number, countexactCovering the number of the accurate negative keywords for the common root, wherein alpha and beta are adjustment coefficients;
and finding the optimal root word combination by using a gradient descent method to serve as a phrase negative keyword set. Wherein, the optimal root word combination can also be found by using genetic algorithm and the like.
S207, eliminating invalid accurate negative keywords in the accurate negative keyword set, and eliminating invalid phrase negative keywords in the phrase negative keyword set. Specifically, the culling rule is: taking the accurate negative keywords in the accurate negative keyword set, which are the same as the account keywords, as invalid accurate negative keywords to be removed; and eliminating the phrase negative keywords contained by the account keywords in the phrase negative keyword set as invalid phrase negative keywords.
S208, if the launching platform has limitation on the number of the added keywords, the high-frequency and high-click search words are preferentially denied.
Wherein, preferentially negating high frequency, high click search terms comprises:
the total number of the accurate negative keywords is recorded as ExactNegNum, the maximum additive value of the accurate negative keywords is recorded as MaxExactNegNum, the total number of the phrase negative keywords is recorded as NegNum, and the maximum additive value of the phrase negative keywords is recorded as MaxNegNum.
If the exact NegNum is larger than the MaxExactNegNum, counting the lengths of all the accurate negative keywords of the account, calculating the average length theta and the standard deviation sigma of the lengths, deleting the accurate negative keywords with the lengths larger than (theta +3 sigma), and updating the exact NegNum.
If the exact NegNum is larger than MaxExactNegNum, counting the number of clicks click of the search words of the account in the past n days, click _ time, removing the accurate negative keywords of which the click _ time is less than the threshold value from the accurate negative keyword set, and updating the ExcctNegNum.
If the exact NegNum is larger than the MaxExactNegNum, counting the occurrence times imexpression _ time of the search words of the past n days of the account, removing the accurate negative keywords of which the imexpression _ time is less than the threshold value from the accurate negative keyword set, and updating the ExcctNegNum.
If the exact negnum is larger than the MaxExactNegNum, the number of the accurate negative keywords to be deleted is DeleteExactNegNum which is exact negNum-MaxExactNegNum, all the accurate negative keywords are sorted according to the length, and the DeleteExactNegNum accurate negative keywords with the longest length are deleted from the accurate negative keyword set.
If NegNum is larger than MaxNegNum, the number of the phrase negative keywords to be deleted is DeleteNegNum which is NegNum-MaxMegNum, all the phrase negative keywords are sorted according to length, and the DeleteNeNum phrase negative keywords with the longest length are deleted from the phrase negative keyword set.
The foregoing only shows a specific embodiment of filtering and filtering negative keywords by using criteria such as click number, frequency, length index, and the like, and optionally, other data indexes such as exposure rate, consumption, and the like may also be used for filtering and filtering.
S209, uploading the accurate negative keyword set and the phrase negative keyword set to a corresponding account of the launching platform. If the algorithm continues to run, returning to S201, if not, ending.
The method for automatically negating the keywords in the keyword advertisement delivery provided by the embodiment of the invention can monitor and analyze the advertisement accounts in real time without interruption so as to automatically mine and add the negative keywords, can perform automatic negative keyword operation on a plurality of accounts in real time, can complete algorithm processing and uploading updating of the automatic negative keywords within seconds, reduces the consumption of manpower and material resources, and improves the operation efficiency. And by utilizing a natural language processing algorithm to analyze texts, accurate negative keywords can be mined, so that the updated negative keywords can play a role in filtering invalid search words, and the invalid consumption of advertisers is reduced.
In a second aspect, an embodiment of the present invention provides an apparatus for automatically rejecting keywords in keyword advertisement delivery, where the apparatus includes:
and the acquisition module is used for automatically acquiring the keywords, the negative keywords and the search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database. A determination module to determine an exact-negative keyword set. And the extraction module is used for extracting the available common root words in the accurate negative keyword set to serve as the phrase negative keyword set. And the eliminating module is used for eliminating invalid accurate negative keywords in the accurate negative keyword set and eliminating invalid phrase negative keywords in the phrase negative keyword set. And the uploading module is used for uploading the accurate negative keyword set and the phrase negative keyword set to the corresponding account of the releasing platform.
The automatic negative keyword device for keyword advertisement delivery, provided by the embodiment of the invention, is used for executing the method, detecting and analyzing the advertisement account in real time without interruption so as to automatically mine and add the negative keywords, and can complete algorithm processing and uploading updating of the automatic negative keywords within seconds, so that the consumption of manpower and material resources is reduced, and the operation efficiency is improved. And accurate negative keywords can be mined, so that the updated negative keywords can have the effect of filtering invalid search words, and the invalid consumption of advertisers is reduced.
It is noted that, in this document, relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. In addition, "front", "rear", "left", "right", "upper" and "lower" in this document are all referred to the placement state shown in the drawings.
Finally, it should be noted that: the above examples are only for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (9)
1. An automated keyword advertisement placement negative keyword method, the method comprising:
automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database;
determining an exact negative keyword set;
extracting available common roots in the accurate negative keyword set to serve as a phrase negative keyword set;
rejecting invalid accurate negative keywords in the accurate negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set;
uploading the accurate negative keyword set and the phrase negative keyword set to a corresponding account of a release platform;
the determining of the set of exact negative keywords comprises:
inquiring whether a negative keyword white list exists or not, and if so, determining search terms except the negative keyword white list as an accurate negative keyword set;
if no negative keyword white list exists, screening out search words with semantic relevance lower than a threshold value through a TF-IDF algorithm, adding the search words to an accurate negative keyword set, screening out search words with negative emotion through a Bayesian classification model, and adding the search words to the accurate negative keyword set.
2. The keyword advertisement placement automated negative keyword method of claim 1, further comprising: and automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time, storing the keywords, the negative keywords and the search word data in a database, and automatically adding binding if a negative keyword universal package is not added.
3. The method of claim 1, wherein the screening search terms with semantic relevance lower than a threshold value by a TF-IDF algorithm to be added to the set of exact negative keywords comprises:
taking the keyword set as a target keyword text, and performing word segmentation by using a hidden Markov model to obtain a keyword root set;
performing word segmentation on the search word text by using a hidden Markov model to obtain a search word root set;
calculating the word frequency TF of the search word root by the following formula Iω，
Calculating the inverse file frequency IDF by the following formula IIω：
Obtaining TF-TDF value TF-IDF of the search term by the following formula three:
and screening out search words with TF-IDF values lower than a threshold value, and adding the search words to the accurate negative keyword set.
4. The method of claim 1, wherein the screening of negative emotion search terms by a bayesian classification model to an exact negative keyword set comprises:
loading an emotion analysis dictionary;
calculating the probability that the search word belongs to the negative emotion through the following formula four:
where P is the probability that the search term belongs to a negative emotion, count (d)iC) search for the word root for negative emotions diNumber of occurrences, VcThe total number of negative emotion root words is n, and the total number of all root words is n;
adding the search words with the negative probability higher than a threshold value to the accurate negative keyword set.
5. The method of claim 1 or 2, wherein the extracting available common roots in the set of exact negative keywords as the set of phrase negative keywords comprises:
taking the accurate negative keyword set as a text, and performing word segmentation by using a hidden Markov model to obtain an accurate negative keyword root set;
extracting common roots in the accurate negative keyword root set, and constructing a target function:
minα·countNeg(x)-β·countexact(x)，
wherein x is a common root combination, countNegTo share the root number, countexactCovering the number of the accurate negative keywords for the common root, wherein alpha and beta are adjustment coefficients;
and finding the optimal root combination by using a gradient descent method to serve as a phrase negative keyword set.
6. The keyword advertisement delivery automated negative keyword method of claim 1 or 2, wherein the rejecting invalid exact negative keywords in the exact negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set comprises:
taking the accurate negative keywords in the accurate negative keyword set, which are the same as the account keywords, as invalid accurate negative keywords to be removed;
and eliminating the phrase negative keywords contained by the account keywords in the phrase negative keyword set as invalid phrase negative keywords.
7. The method for automatically rejecting the negative keywords in the keyword advertisement delivery according to claim 1 or 2, further comprising rejecting the invalid precise negative keywords in the set of precise negative keywords and rejecting the invalid phrase negative keywords in the set of phrase negative keywords, and preferentially rejecting the high-frequency high-click search terms if the delivery platform has a limit on the number of the negative keywords that can be added.
8. The keyword advertisement placement automated negative keyword method of claim 7, wherein the preferentially negating high frequency, high click search terms comprises:
counting the total number of the accurate negative keywords as ExactNegNum, wherein the maximum addable value of the accurate negative keywords is MaxExactNegNum, the total number of the phrase negative keywords is NegNum, and the maximum addable value of the phrase negative keywords is MaxNegNum;
if the exact NegNum is larger than the MaxExactNegNum, counting the length of each accurate negative keyword of the account, calculating the average length theta and the standard deviation sigma of the length, deleting the accurate negative keywords with the length larger than (theta +3 sigma), and updating the exact NegNum;
if the exact NegNum is larger than the MaxExactNegNum, counting the click times click _ time of the search words of the account in the past n days, removing the accurate negative keywords of which the click _ time is less than the threshold value from the accurate negative keyword set, and updating the exact NegNum;
if the exact NegNum is larger than the MaxExactNegNum, counting the occurrence times impresson _ time of the search words of the past n days of the account, removing the accurate negative keywords of which the impresson _ time is less than the threshold value from the accurate negative keyword set, and updating the exact NegNum;
if the exact NegNum is more than MaxExactNegNum, the number of the accurate negative keywords to be deleted is DeleteExactNegNum which is exact NegNum-MaxExactNegNum, all the accurate negative keywords are sorted according to the length, and the DeleteExactNegNum accurate negative keywords with the longest length are deleted from the accurate negative keyword set;
if NegNum is larger than MaxNegNum, the number of the phrase negative keywords to be deleted is DeleteNegNum which is NegNum-MaxMegNum, all the phrase negative keywords are sorted according to length, and the DeleteNegNum phrase negative keywords with the longest length are deleted from the phrase negative keyword set.
9. An automated keyword advertisement delivery negative keyword apparatus, characterized in that the apparatus is controlled by a method according to any one of claims 1 to 8, the apparatus comprising:
the acquisition module is used for automatically acquiring keywords, negative keywords and search word data of the advertisement account in real time and storing the keywords, the negative keywords and the search word data in a database;
a determination module for determining a set of exact negative keywords;
the extraction module is used for extracting an available common root word in the accurate negative keyword set to serve as a phrase negative keyword set;
the rejecting module is used for rejecting invalid accurate negative keywords in the accurate negative keyword set and rejecting invalid phrase negative keywords in the phrase negative keyword set;
and the uploading module is used for uploading the accurate negative keyword set and the phrase negative keyword set to the corresponding account of the releasing platform.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910962210.1A CN110717104B (en) | 2019-10-11 | 2019-10-11 | Keyword advertisement putting automatic negative keyword method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910962210.1A CN110717104B (en) | 2019-10-11 | 2019-10-11 | Keyword advertisement putting automatic negative keyword method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110717104A CN110717104A (en) | 2020-01-21 |
| CN110717104B true CN110717104B (en) | 2022-05-20 |
Family
ID=69211411
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910962210.1A Active CN110717104B (en) | 2019-10-11 | 2019-10-11 | Keyword advertisement putting automatic negative keyword method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110717104B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112015994B (en) * | 2020-09-09 | 2023-09-15 | 平安科技（深圳）有限公司 | Drug recommendation method, device, equipment and medium |
| CN113032683B (en) * | 2021-04-28 | 2021-12-24 | 玉米社(深圳)网络科技有限公司 | Method for quickly segmenting words in network popularization |
| 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 |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101496403A (en) * | 2006-07-31 | 2009-07-29 | 微软公司 | Associating advertisements with on-demand media content |
| CN103164454A (en) * | 2011-12-15 | 2013-06-19 | 百度在线网络技术（北京）有限公司 | Keyword grouping method and keyword grouping system |
| CN103425705A (en) * | 2012-05-24 | 2013-12-04 | 阿里巴巴集团控股有限公司 | Negative keyword acquiring method and device and search method and device |
| CN103473233A (en) * | 2012-06-06 | 2013-12-25 | 百度在线网络技术（北京）有限公司 | Method and device for determining optimization information of search promotion units |
| CN104077290A (en) * | 2013-03-26 | 2014-10-01 | 腾讯科技（深圳）有限公司 | Method and device for generating promoted accounts |
| CN104239586A (en) * | 2014-10-16 | 2014-12-24 | 北京奇虎科技有限公司 | Method and device for processing information material file |
| CN105138690A (en) * | 2015-09-18 | 2015-12-09 | 北京博雅立方科技有限公司 | Method and device for determining keywords |
| CN106570019A (en) * | 2015-10-10 | 2017-04-19 | 北京国双科技有限公司 | Key word configuration method and device |
| CN108255881A (en) * | 2016-12-29 | 2018-07-06 | 北京国双科技有限公司 | It is a kind of to generate the method and device for launching keyword |
| WO2019122990A1 (en) * | 2017-12-23 | 2019-06-27 | Yogesh Chunilal Rathod | Providing interface for identifying one or more types of user actions & criteria specific users |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102004979A (en) * | 2009-09-03 | 2011-04-06 | 叶克 | System and method for providing commodity matching and promoting services |
| US10482146B2 (en) * | 2016-05-10 | 2019-11-19 | Massachusetts Institute Of Technology | Systems and methods for automatic customization of content filtering |
-
2019
- 2019-10-11 CN CN201910962210.1A patent/CN110717104B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101496403A (en) * | 2006-07-31 | 2009-07-29 | 微软公司 | Associating advertisements with on-demand media content |
| CN103164454A (en) * | 2011-12-15 | 2013-06-19 | 百度在线网络技术（北京）有限公司 | Keyword grouping method and keyword grouping system |
| CN103425705A (en) * | 2012-05-24 | 2013-12-04 | 阿里巴巴集团控股有限公司 | Negative keyword acquiring method and device and search method and device |
| CN103473233A (en) * | 2012-06-06 | 2013-12-25 | 百度在线网络技术（北京）有限公司 | Method and device for determining optimization information of search promotion units |
| CN104077290A (en) * | 2013-03-26 | 2014-10-01 | 腾讯科技（深圳）有限公司 | Method and device for generating promoted accounts |
| CN104239586A (en) * | 2014-10-16 | 2014-12-24 | 北京奇虎科技有限公司 | Method and device for processing information material file |
| CN105138690A (en) * | 2015-09-18 | 2015-12-09 | 北京博雅立方科技有限公司 | Method and device for determining keywords |
| CN106570019A (en) * | 2015-10-10 | 2017-04-19 | 北京国双科技有限公司 | Key word configuration method and device |
| CN108255881A (en) * | 2016-12-29 | 2018-07-06 | 北京国双科技有限公司 | It is a kind of to generate the method and device for launching keyword |
| WO2019122990A1 (en) * | 2017-12-23 | 2019-06-27 | Yogesh Chunilal Rathod | Providing interface for identifying one or more types of user actions & criteria specific users |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110717104A (en) | 2020-01-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110717104B (en) | Keyword advertisement putting automatic negative keyword method and device | |
| US11507975B2 (en) | Information processing method and apparatus | |
| CN109190017B (en) | Method and device for determining hotspot information, server and storage medium | |
| TWI654530B (en) | Method and device for screening and promoting keywords | |
| CN110874530B (en) | Keyword extraction method, keyword extraction device, terminal equipment and storage medium | |
| US20180260385A1 (en) | Symbol management | |
| WO2018040068A1 (en) | Knowledge graph-based semantic analysis system and method | |
| CN110263248A (en) | A kind of information-pushing method, device, storage medium and server | |
| CN105389349A (en) | Dictionary updating method and apparatus | |
| CN109635082A (en) | Policy implication analysis method, device, computer equipment and storage medium | |
| CN109815499B (en) | Information association method and system | |
| CN112287682B (en) | Method, device and equipment for extracting subject term and storage medium | |
| CN108829661A (en) | A kind of subject of news title extracting method based on fuzzy matching | |
| CN111047092A (en) | Dispute case victory rate prediction method and device, computer equipment and storage medium | |
| CN105068986A (en) | Method for filtering comment spam based on bidirectional iteration and automatically constructed and updated corpus | |
| CN110389963A (en) | The recognition methods of channel effect, device, equipment and storage medium based on big data | |
| CN109508557A (en) | A kind of file path keyword recognition method of association user privacy | |
| CN102103604B (en) | Method and device for determining core weight of term | |
| CN112989235A (en) | Knowledge base-based internal link construction method, device, equipment and storage medium | |
| CN110489759B (en) | Text feature weighting and short text similarity calculation method, system and medium based on word frequency | |
| CN110991169A (en) | Method and device for identifying risk content variety and electronic equipment | |
| CN108427769B (en) | Character interest tag extraction method based on social network | |
| CN108921431A (en) | Government and enterprise customers clustering method and device | |
| CN112069388B (en) | Entity recommendation method, system, computer device and computer readable storage medium | |
| CN111736804B (en) | Method and device for identifying App key function based on user comment |
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 |