A method and system for automatically adding or denying words in bidding advertisements

Abstract

The present invention relates to a method and system for automatically adding or rejecting words in bidding advertisements, the method comprising the following steps: obtaining a target search word through at least one advertising platform; comparing the similarity between the target search word and the keyword in the account, and sending the comparison result to a client; when the target search word meets a preset adding condition, adding the target search word to the account of the target keyword; the adding condition at least includes that the similarity between the target search word and the target keyword is greater than the adding threshold. The function of the present invention is to realize the automatic processing of adding good words to the search words that are matched every day and do not exist in the account according to the adding or rejecting rule.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet advertising, and more specifically, to a method and system for automatically adding and rejecting words in bidding advertisements. Background Art

[0002] Bidding advertising is a process where advertisers set keywords to be placed and bid on them to compete for keyword ad rankings. Almost all advertising promotion platforms provide advertisers with account management structures such as promotion plans and promotion units to facilitate advertisers in managing keywords.

[0003] However, in order to expand the amount of delivery, advertising platforms generally expand the keywords. For example, if an advertiser places the keyword "car", in addition to matching the car itself, it will also match its related words and similar words, such as "family car", "suv car", "car sales list", "which car is the best" and other search terms.

[0004] Advertising platforms generally provide advertisers with reports on all search terms and their advertising data that match each keyword, allowing advertisers to view and process search terms that are "matched but do not exist in the account." Each search term that is matched but does not exist in the account may be matched for a long time and generate long-term advertising consumption, affecting the long-term effects of the account. If a good term is matched, it may bring better advertising effects to the advertiser, and the term can be added to the account. However, if a bad term is matched, it will bring worse effects to the advertiser, and the term can be added to the negative term library to avoid matching. This requires advertisers to determine whether to add good terms or deny bad terms through analysis reports.

[0005] However, the reports are lengthy and the search terms are numerous and complex, requiring a large amount of human resources to analyze. This is a huge cost investment and has low work efficiency. Especially for medium and large advertisers, it is generally difficult for a staff member to process all the search terms for the day even if he or she uses a whole day's work.

[0006] In the prior art, a Chinese invention patent with patent number 201810800654.0 discloses a method and device for intelligent keyword advertising delivery based on big data technology, which processes the acquired delivery data based on artificial intelligence technology and automatically adjusts prices and executes advertising delivery based on bidding ranking to achieve adjustments to keywords that need to be adjusted in accordance with the delivery strategy.

[0007] However, the above technical solution has the following problems: since it is based on automatic price adjustment based on paid ranking, and since the paid ranking is personalized through ocpc (different users get different rankings) and inaccurate, the above technology cannot be universally used for wide application. Summary of the invention

[0008] In order to solve the above problems, the present invention provides a method and system for automatically adding and removing bad words in bidding advertisements to realize the automatic processing of adding good words. The technical solution is as follows:

[0009] A method for automatically adding or removing words in a bidding advertisement, comprising:

[0010] Obtain target search terms through at least one advertising platform;

[0011] The target search term is compared with the key words in the account for similarity, and the comparison result is sent to the client.

[0012] Furthermore, it also includes:

[0013] When the target search term meets the preset adding condition, the target search term is added to the target keyword account;

[0014] The word adding condition at least includes that the similarity between the target search word and the target keyword is greater than a word adding threshold.

[0015] Furthermore, when the target search term satisfies a preset adding condition, adding the target search term to the account includes:

[0016] Convert the target search term into the target element term according to the matching rules;

[0017] Converting keywords into account element words according to the matching rules;

[0018] The word adding condition at least includes that the similarity between the target element word and the account element word is greater than the word adding threshold;

[0019] The calculation method of the similarity a is:

[0020] a=x / y

[0021] Where x is the number of elements in the account element word that are identical to the target element word, and y is the number of elements in the target element word;

[0022] Acquire the matched account element word according to the word addition condition;

[0023] When the matched account element word is one, adding the target search word to the target keyword account corresponding to the account element word;

[0024] When more than one account element word is matched, the similarities of the matched account element words are compared, the one with the highest similarity is selected as the target keyword, and the target search word is added to the target keyword account;

[0025] When there are more than one account element words with the highest similarity, the advertisement data of the account element words are compared, the highest weight in the advertisement data is selected as the target keyword, and the target search word is added to the target keyword account.

[0026] Furthermore, the matching rules include:

[0027] Establish root unit table and element unit table;

[0028] The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap;

[0029] The element unit table includes N element modules, the element modules include numbers, letters and / or symbols, and different element modules have different contents. The element modules correspond to the root modules one by one and do not overlap.

[0030] Segment the target search term and / or the keyword in the account to obtain word roots;

[0031] The root word is converted into a target element word and / or an account element word according to a one-to-one correspondence between the root word unit table and the element unit table.

[0032] Furthermore, it also includes:

[0033] When the target search term meets the preset negative word condition, the target search term is added to the account negative word library of the target keyword;

[0034] The negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword.

[0035] Furthermore, it also includes:

[0036] When at least one of the word roots obtained by the segmentation of the target search term cannot be converted into an element, negative word information is generated and sent to the client, and the word root that cannot be converted into an element is added or not added to the account negative word library according to the negative word feedback instruction of the client.

[0037] Furthermore, the advertising data includes consumption, clicks, impressions, conversions, orders, inquiries, payments and payment amounts, and the weights are consumption, clicks, impressions, conversions, orders, inquiries, payments or payment amounts.

[0038] Furthermore, it also includes: when the target search word meets the preset word addition condition, generating word addition information and sending it to the client, and adding the target search word to the target keyword account or not adding it according to the word addition feedback instruction of the client;

[0039] The word-added feedback instruction includes a word-added instruction or a no-word-added instruction.

[0040] Furthermore, it also includes: when the target search word meets the preset negative word condition, generating negative word information and sending it to the client, and adding the target search word to the account negative word library of the target keyword or not adding it according to the negative word feedback instruction of the client;

[0041] The negative word feedback instruction includes a negative word adding instruction and a negative word not adding instruction.

[0042] Further, the comparing the similarity between the target search term and the keywords in the account and sending the comparison result to the client includes: comparing the similarity between the target search term and the keywords in the account, taking the keyword with the highest similarity as the target keyword, generating word addition information and sending it to the client, and adding the target search term to the account of the target keyword or not according to the word addition feedback instruction of the client;

[0043] The word-added feedback instruction includes a word-added instruction or a no-word-added instruction.

[0044] Furthermore, when there is only one keyword with the highest similarity, the keyword with the highest similarity is used as the target keyword, and the added word information is generated and sent to the client;

[0045] When there are more than one keywords with the highest similarity, the advertisement data of the keywords are compared, the keyword with the highest weight in the advertisement data is selected as the target keyword, and the added word information is generated and sent to the client.

[0046] Furthermore, the advertising data includes consumption, clicks, impressions, conversions, orders, inquiries, payments and payment amounts, and the weights are consumption, clicks, impressions, conversions, orders, inquiries, payments or payment amounts.

[0047] According to another aspect of the present invention, there is also provided a bidding advertisement automatic word addition and word rejection system, comprising:

[0048] A target search term acquisition unit, configured to acquire a target search term from at least one advertising platform;

[0049] A word adding judgment unit, configured to add the target search word to the target keyword account when the target search word meets a preset word adding condition;

[0050] The word addition condition at least includes that the similarity between the target search word and the target keyword is greater than the word addition threshold; and

[0051] a negative word judgment unit, configured to add the target search word to the account negative word library of the target keyword when the target search word meets a preset negative word condition;

[0052] The negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword.

[0053] Furthermore, it also includes:

[0054] A matching unit, configured to convert a target search term into a target element term according to a matching rule; and / or

[0055] Converting keywords into account element words according to the matching rules;

[0056] The matching rules include:

[0057] Establish root unit table and element unit table;

[0058] The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap;

[0059] The element unit table includes N element modules, the element modules include numbers, letters and / or symbols, and different element modules have different contents. The element modules correspond to the root modules one by one and do not overlap.

[0060] Segment the target search term and / or the keyword in the account to obtain word roots;

[0061] The root word is converted into a target element word and / or an account element word according to a one-to-one correspondence between the root word unit table and the element unit table.

[0062] Furthermore,

[0063] The word adding condition at least includes that the similarity between the target element word and the account element word is greater than the word adding threshold;

[0064] The calculation method of the similarity a is:

[0065] a=x / y

[0066] Where x is the number of elements in the account element word that are identical to the target element word, and y is the number of elements in the target element word.

[0067] The method and system for automatically adding and rejecting words in bidding advertisements of the present invention reduces the possibility of human misjudgment and greatly improves the accuracy of processing; on the other hand, it changes the mode of processing one by one into batch processing, thereby improving work efficiency.

[0068] Further features and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments of the present invention with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present invention and, together with the description, serve to explain the principles of the present invention.

[0070] Figure 1 This is a method logic diagram of the method for automatically adding or removing words from bidding advertisements according to Embodiment 1 of the present invention;

[0071] Figure 2 This is a matching flow chart of the method for automatically adding or removing words from a bidding advertisement in Embodiment 1 of the present invention;

[0072] Figure 3 This is a method logic diagram of the method for automatically adding or removing words from bidding advertisements according to Embodiment 2 of the present invention;

[0073] Figure 4 This is a flow chart of a word adding method in a method for automatically adding or rejecting words in a bidding advertisement according to Embodiment 2 of the present invention;

[0074] Figure 5 This is a structural block diagram of the system for automatically adding and removing words from bidding advertisements according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0075] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless otherwise specifically stated.

[0076] Techniques and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the techniques and equipment should be considered part of the specification.

[0077] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0078] Example 1

[0079] An embodiment of the present invention discloses a method for automatically adding or rejecting words in a bidding advertisement. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0080] Figure 1 The following is a logic diagram of a method for automatically adding or removing words from a bidding advertisement according to an embodiment of the present invention. As shown in the figure, the method includes the following steps:

[0081] Step S100: Acquire a target search term through at least one advertising platform.

[0082] In order to promote their products, advertisers generally place advertisements on multiple advertising platforms, such as Baidu, 360, Shenma, Taobao, Douyin, Kuaishou and other platforms. In the embodiments of the present invention, they are all referred to as advertising platforms. The advertising platforms provide users with a search function. When the user enters a search term, due to the keyword expansion function provided by the platform, multiple advertisements related to the search term will be displayed based on the search term. They may be the same, similar, related, or even unrelated.

[0083] After an advertiser places a keyword and a user triggers the keyword through a search term, the advertising platform will generally provide the advertiser with a report on all search terms and advertising data that match each keyword, allowing the advertiser to view and process the matched search terms that do not exist in the account.

[0084] In an embodiment of the present invention, by utilizing API (application programming interface) technology, the API interface provided by the corresponding advertising platform can be used to automatically obtain all data reports provided by each advertising platform for advertisers, and extract the matched target search terms that do not exist in the account.

[0085] In another embodiment of the present invention, the target search term may also be obtained from the advertiser's client. By having the advertiser upload the target search term independently, a more comprehensive target search term basis may be obtained.

[0086] For example, the keyword placed by the advertiser is "data recovery tool", and the target search terms obtained from the advertising platform are "data recovery software", "data", "restore data", etc.

[0087] Step S210: Convert the target search word into a target element word according to the matching rule.

[0088] Figure 2 This is a matching flow chart of an embodiment of the present invention. As shown in the figure, the matching rules are:

[0089] Step S211, establishing a root unit table and an element unit table;

[0090] The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap, and N is an integer greater than 1.

[0091] For example, the root word module "dad" contains roots with the same or similar meanings as "dad", so roots such as father, father, dad, daddy, biological father, and adoptive father all belong to this root word module. Only a few are listed here for illustration, not all.

[0092] The element unit table includes N element modules. Each element module contains numbers, letters, or / and symbols, and the content of different element modules is different. Each element module corresponds one-to-one with a root word module and they do not overlap.

[0093] In an embodiment of the present invention, the element module contains numbers. As shown in the following table, element module 1 corresponds to the root word module of "dad", element module 2 corresponds to the root word module of "mom", and element module 3 corresponds to the root word module of "father-in-law". Only the above 3 are listed here for illustration, not a comprehensive list.

[0094]

[0095]

[0096] Step S212: Segment the target search term to obtain root words.

[0097] By using a word segmentation tool to segment the target search term to obtain root words, the target search term "data recovery software" can be segmented into three root words: data, recovery, and software.

[0098] Step S213: Convert the root words into target element words according to the one-to-one correspondence between the root word unit table and the element unit table.

[0099] Then the target search term "data recovery software" can be converted into the target element word: 4_5_6. For the convenience of distinction, each element is separated by an interval symbol _. In other embodiments of the present invention, the interval symbol can also be a space, a letter, a punctuation mark, etc.

[0100] In fact, sometimes the search term not only contains root words, but also contains symbols, spaces, etc. In addition, some root words, such as "de" (的 in Chinese), generally have no specific meaning and can be directly ignored during matching. Therefore, in an embodiment of the present invention, a module of meaningless root words is also set. This module includes root words, punctuation marks, and symbols that are not controversial, and there is no corresponding element module for this module, or the corresponding element module of this module is empty. Therefore, when the segmented root word contains the content included in the above module of meaningless root words, there will be no corresponding element for this root word.

[0101] For example, if the target search term is "data recovery de software", according to the matching rule, it can be converted into the target element word: 4_5_6.

[0102] For the convenience of clustering and comparing different elements, in an embodiment of the present invention, the order of appearance of elements in the element word combination is not distinguished. For example, "data recovery software" can be converted into the account element word: 4_5_6, and "recovery data software" can be converted into the element word: 5_4_6. Since the order of elements is not distinguished, the element words 4_5_6 and 5_4_6 are the same.

[0103] In fact, some word roots may not be matched to elements. The reason may be that the advertiser does not need the word root, or the word root library may not be comprehensive enough. For example, if the target search term is "data recovery courseware", if the word "courseware" has no corresponding element in the element table, then the target search term can only be converted to: 5_4_courseware.

[0104] In an embodiment of the present invention, "courseware" can be called a brand new word. At this time, the corresponding information of the brand new word needs to be generated into negative word information and sent to the client. The advertiser of the client decides whether to add it to the negative word library. If the advertiser decides to add it to the negative word library, the client will feedback an adding instruction to add the brand new word to the negative word library of the planning unit of the matched keyword.

[0105] For example, if the keyword placed by the advertiser is: data recovery software, and the search term "data recovery courseware" on the advertising platform matches the search term, "data recovery courseware" is first converted into a target element word. Since "courseware" has no corresponding element in the element table, it is converted into 5_4_courseware. Then, the corresponding information of the new word "courseware" and the keyword data recovery software is used to generate negative word information and send it to the client. The advertiser of the client decides whether to add it to the negative word library.

[0106] Step S220: Convert the keywords into account element words according to the matching rules.

[0107] In an embodiment of the present invention, for example, the keyword "data recovery tool" can be converted into the account element words: 4_5_6.

[0108] An account is a tool provided by an advertising platform to advertisers that can be used to browse, query or set keyword information. Depending on the advertising platform, the content and module units contained in the account may vary, but basic keyword information is generally available. Here, the account provided by Baidu is used as an example to illustrate that the advertiser's account includes several plans, each plan contains several units, and each unit includes the keywords placed by the advertiser and a keyword library.

[0109] Step S230, compare the similarity between the target element word and the account element word, take the account element word with the highest similarity as the target keyword, generate word addition information and send it to the client, and add the target search word to the account of the target keyword or not according to the client's word addition feedback instruction, wherein the word addition feedback instruction includes a word addition instruction or a word non-addition instruction.

[0110] Specifically, in an embodiment of the present invention, when there is only one account element word matched, the information of the target keyword is used to generate added word information, which includes a recommended bid, a recommended matching method, a recommended account plan unit to be added, etc., and the added word information is sent to the client; after the client where the advertiser is located receives the added word information, it adds or does not add the target search term to the account plan unit of the target keyword according to the added word instruction or non-added word instruction fed back by the advertiser.

[0111] For example, the target element word is 4_5_6, and the only account element word with the highest similarity to 4_5_6 is 5_4_6. The corresponding keyword is "data recovery software". Then, information about the keyword "data recovery software" is extracted, including recommended bids, recommended matching methods, and recommended account plan units to be added, etc., to generate word addition information and send it to the client. The advertiser decides whether to add the word based on its own needs.

[0112] When there are more than one account element words with the highest similarity, the advertisement data of the account element words are compared, and the one with the highest weight in the advertisement data is selected as the target keyword.

[0113] For example, let's take two account element words with the highest similarity as an example. When the words with the highest similarity to the target search words are "data recovery tools" and "data restore software", it is necessary to compare the advertising data of these two keywords. The advertising data is used to show the value of the keywords.

[0114] In an embodiment of the present invention, API (application programming interface) technology is also used to automatically obtain all keyword data reports provided by each advertising platform from the API interface provided by the corresponding advertising platform, and extract advertising data of all keywords for advertisers.

[0115] The content of the advertising data can be set independently or preset according to the needs of the advertiser. The advertising data includes promotion data, conversion data and transaction data. The promotion data includes consumption, click-through rate, display rate, upper-level consumption, upper-level click-through rate, upper-level display rate, etc. The conversion data includes consultation, phone calls, forms, WeChat, downloads, installations, favorites, attention, adding to shopping carts, etc. The transaction data includes order volume, payment volume, payment amount, transaction volume, revenue, etc. In fact, depending on the advertising platform, the content of the above advertising data may also be different. Therefore, only enumeration is made here for explanation, not exhaustive examples. As long as the data is provided by the advertising platform or advertiser and can reflect the value of keywords, they are all advertising data of the embodiments of the present invention.

[0116] In an embodiment of the present invention, the advertising data includes consumption, click volume, display volume, conversion volume, order volume, consultation volume, payment volume and payment amount.

[0117] The weight is the content in the advertising data that the advertiser values ​​more, which may be one or more items. Therefore, the weight can also be set independently or in advance according to the needs of the advertiser. In the embodiment of the present invention, the weight is consumption.

[0118] Therefore, in an embodiment of the present invention, it is necessary to compare the consumption of the two words "data recovery tool" and "data restoration software", select the one with the highest consumption as the target keyword, generate added word information based on the information of the target keyword, and the added word information includes a recommended bid, a recommended matching method, a recommended account plan unit to be added, etc., and send the added word information to the client; after the client where the advertiser is located receives the added word information, it adds or does not add the target search term to the account plan unit of the target keyword according to the added word instruction or non-added word instruction fed back by the advertiser.

[0119] Furthermore, when there is more than one keyword with the highest weight, determine whether the plan units of each keyword are repeated. If there are a keywords with non-repeated plan units, these a keywords are all used as target keywords, and their information is generated into added word information and recommended to the client. After the advertiser's client receives the added word information, it adds or does not add the target search term to the account plan unit of the target keyword according to the added word instruction or non-added word instruction fed back by the advertiser.

[0120] In addition, the method of this embodiment also provides a method for adding negative words, specifically: when the target search word meets the preset negative word condition, the target search word is added to the account negative word library of the target keyword; the negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword.

[0121] When a keyword with a high bid matches a keyword with a low bid, the average click price of the same search term as the keyword with a low bid in the search term report exceeds its bid. In the embodiment of the present invention, such a term may be referred to as a "high-matching-low" search term. The details will be described in detail in Embodiment 2.

[0122] Example 2

[0123] The embodiment of the present invention discloses a method for automatically adding or removing words in a bidding advertisement, such as Figure 3 As shown, the following steps are included:

[0124] Step S100: Acquire a target search term through at least one advertising platform.

[0125] In order to promote their products, advertisers generally place advertisements on multiple advertising platforms, such as Baidu, 360, Shenma, Taobao, Douyin, Kuaishou and other platforms. In the embodiments of the present invention, they are all referred to as advertising platforms. The advertising platforms provide users with a search function. When the user enters a search term, due to the keyword expansion function provided by the platform, multiple advertisements related to the search term will be displayed based on the search term. They may be the same, similar, or related.

[0126] After an advertiser places a keyword and a user triggers the keyword through a search term, the advertising platform will generally provide the advertiser with a report on all search terms and advertising data that match each keyword, allowing the advertiser to view and process the matched search terms that do not exist in the account.

[0127] In an embodiment of the present invention, by utilizing API (application programming interface) technology, the API interface provided by the corresponding advertising platform can be used to automatically obtain all data reports provided by each advertising platform for advertisers, and extract the matched target search terms that do not exist in the account.

[0128] In another embodiment of the present invention, the target search term can also be obtained from the advertiser's client. By obtaining the target search term through the advertiser's independent uploading, a more comprehensive target search term basis can be obtained.

[0129] For example, the keyword placed by the advertiser is "data recovery tool", and the target search term obtained from the advertising platform is "data recovery software", etc.

[0130] Step S200: When the target search term meets the preset adding condition, the target search term is added to the target keyword account;

[0131] The above-mentioned word addition condition at least includes that the similarity between the target search word and the target keyword is greater than the word addition threshold.

[0132] In an embodiment of the present invention, the similarity a of the similarity is calculated as follows:

[0133] a=x / y

[0134] Where x is the number of elements in the account element word that are identical to the target element word, and y is the number of elements in the target element word.

[0135] Generally speaking, whether to add a target search term is determined by the advertiser's own choice. However, when the target search terms are huge, manual selection is not only inefficient but also prone to errors during the working process.

[0136] In an embodiment of the present invention, a condition for adding words is preset, and the condition for adding words is that the similarity between the target search word and the target keyword is greater than the threshold for adding words. The greater the similarity, the higher the relevance to the keyword, which means the higher its value. Therefore, it needs to be added to the account. When the user of the advertising platform uses the target search word to search, the advertiser's advertisement is more likely to be displayed and it is easier to obtain a higher advertisement ranking. The specific value of the threshold for adding words can be set by the advertiser independently based on experience or obtained through simulation experiments.

[0137] When the target search term meets the conditions for adding a term, a command for adding a term is sent to the account. Here, Baidu search platform is used as an example. The account structure of an advertiser generally includes several plans and several units under the plan. The account structure provided by other advertising platforms for advertisers may include multiple units. In short, the account structure is basically similar. According to the location of the target keyword, the target search term is added to the account plan unit corresponding to the target keyword.

[0138] In order to save time and obtain similarity more quickly, in the embodiment of the present invention, the specific word adding method is as follows: Figure 4 As shown, including:

[0139] Step S210: Convert the target search word into a target element word according to the matching rule.

[0140] The matching rules are:

[0141] Step S211, establishing a root unit table and an element unit table;

[0142] The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap, and N is an integer greater than 1.

[0143] For example, the root word module "dad" contains roots with the same or similar meanings as "dad", so roots such as father, father, dad, daddy, biological father, and adoptive father all belong to this root word module. Only a few are listed here for illustration, not all.

[0144] The element unit table includes N element modules, which include numbers, letters and / or symbols. Different element modules have different contents. The element modules correspond to the root modules one by one and do not overlap.

[0145] In an embodiment of the present invention, the element module includes numbers. As shown in the table in Example 1, element module 1 corresponds to the root module of the father, element module 2 corresponds to the root module of the mother, element module 3 corresponds to the root module of the father-in-law, and so on. The technical solution of this part is similar to that of Example 1, so it will not be repeated here.

[0146] Step S212: Segment the target search term to obtain word roots.

[0147] Then the target search term "data recovery software" can be segmented into three root words: data, recovery, and software; the keyword "data recovery tool" can be segmented into data, recovery, and tool.

[0148] Step S213: Convert the root words into target element words according to the one-to-one correspondence between the root word unit table and the element unit table.

[0149] Then the target search term "data recovery software" can be converted into the target element word: 4_5_6. For the convenience of distinction, each element is separated by an interval symbol _. In other embodiments of the present invention, the interval symbol can also be a space, a letter, a punctuation mark, etc.

[0150] In fact, search terms sometimes not only contain root words, but also symbols, spaces, etc. In addition, some root words, such as "de" (的 in Chinese), generally have no specific meaning and can be directly ignored during matching. Therefore, in the embodiments of the present invention, a module for meaningless root words is also set. This module includes undisputed root words, punctuation marks, and symbols, etc., and there is no corresponding element module for this module, or the corresponding element module for this module is empty. Therefore, when the segmented root words contain the content included in the above-mentioned module for meaningless root words, there will be no corresponding element for this root word.

[0151] For example, if the target search term is "data recovery software", according to the matching rule, it can be converted into the target element word: 4_5_6.

[0152] For the convenience of clustering and comparing different elements, in the embodiments of the present invention, the order of appearance of elements in the element word combination is not distinguished. For example, "data recovery software" can be converted into the account element word: 4_5_6, and "recovery data software" can be converted into the element word: 5_4_6. Since the order of elements is not distinguished, the element words 4_5_6 and 5_4_6 are the same.

[0153] Step S220: Convert the keyword into an account element word according to the matching rule.

[0154] In the embodiments of the present invention, for example, the keyword "data recovery tool" can be converted into the account element word: 4_5_6.

[0155] Step S230: Obtain the matched account element words according to the word addition condition.

[0156] In the embodiments of the present invention, taking the word addition threshold of 0.5 as an example for illustration. That is, all account element words with a similarity greater than 0.5 to the target element word can be matched.

[0157] In another embodiment of the present invention, the threshold for adding words is set by the advertiser independently, so that the advertiser can judge whether to adjust the threshold according to actual needs, so as to further improve the revenue and efficiency for the advertiser.

[0158] Since different advertisers use different expansion levels, some advertisers have a small expansion range, while others have a large expansion range. For advertisers with a small expansion range, most of the search terms that are matched by the element need to be added to the account, and the threshold for adding words should be set to a smaller value. Therefore, in another embodiment of the present invention, the threshold for adding words can also be pre-set by the advertiser according to different elements, for example, the threshold for adding words matched to element 4 is 0.1, the threshold for adding words matched to element 5 is 0.5, and the threshold for adding words matched to element 6 is 0.3.

[0159] Step S241: When there is only one account element word matched, add the target search word to the target keyword account corresponding to the account element word;

[0160] That is, when there is only one account element word converted from the keywords in the account and the target element word with a similarity greater than 0.5, for example, the keywords are "data recovery", "data tool", "data search tool", "data recovery tool" or "data restoration tool", that is, the account element words are 4_5, 4_6, 1_4_6, 4_5_6, 4_5_6, among which, the familiarity between 4_5, 4_6, 1_4_6 and 4_5_6 is 2 / 3, equal to 0.67, which is greater than the word addition threshold of 0.5, and the familiarity between 4_5_6 and 4_5_6 is 3 / 3, equal to 1, which is greater than the word addition threshold of 0.5, then the target search term "data recovery software" is directly added to the corresponding target keyword account, and according to the different plans and units in which the target keyword is located in the account, it needs to be added to the corresponding plan unit, and the bid of the target keyword is adopted.

[0161] Step S242: When there are more than one matched account element words, compare the similarities of the matched account element words, select the one with the highest similarity as the target keyword, and add the target search word to the target keyword account;

[0162] In an embodiment of the present invention, four account element words are matched: 4_5, 4_6, 1_4_6 and 4_5_6. The four account element words correspond to four keywords: "data recovery", "data tool", "data search tool" and "data recovery tool". For example, it is necessary to compare the similarity between each keyword and the target search word. Among the above four keywords, the one with the highest similarity to the target search word is "data recovery tool", and its similarity is 1. At this time, the target search word "data recovery software" is added to the account plan unit corresponding to "data recovery tool", and the bid of "data recovery tool" is used as the bid of "data recovery software".

[0163] Step S243: When there are more than one account element words with the highest similarity, compare the advertising data of the account element words, select the one with the highest weight in the advertising data as the target keyword, and add the target search word to the target keyword account.

[0164] For example, there are 5 matched account element words: 4_5, 4_6, 1_4_6, 4_5_6 and 5_4_6. The 5 account element words correspond to 5 keywords: "data recovery", "data tool", "data search tool", "data recovery tool", "data restore software". The words with the highest similarity to the target search words are "data recovery tool" and "data restore software", and their similarities are both 1. At this time, it is necessary to compare the advertising data of these two keywords. The advertising data is used to show the value of the keywords.

[0165] In an embodiment of the present invention, API (application programming interface) technology is also used to automatically obtain all keyword data reports provided by each advertising platform from the API interface provided by the corresponding advertising platform, and extract advertising data of all keywords for advertisers.

[0166] The content of the advertisement data can be set independently or preset according to the needs of the advertiser. In the embodiment of the present invention, the advertisement data is the same as the advertisement data of Embodiment 1 and will not be described in detail here.

[0167] The weight is the content in the advertising data that the advertiser values ​​more, which may be one or more items. Therefore, the weight can also be set independently or in advance according to the needs of the advertiser. In the embodiment of the present invention, the weight is consumption.

[0168] Therefore, in the embodiment of the present invention, it is necessary to compare the consumption of the two words "data recovery tool" and "data restoration software", and add the target search word to the account plan unit of the keyword with the highest consumption.

[0169] Furthermore, the method may include step S244: when there is more than one keyword with the highest weight, determining whether the plan units of the keywords are repeated; if there are a plan units with non-repeated keywords, adding the target search term to the plan unit where the a keywords are located.

[0170] In order to make the addition of words controllable and decide whether to add words according to the needs of the advertiser, the following steps are added in the embodiment of the present invention:

[0171] Step S300: When the target search term meets the preset adding condition, the adding information is generated and sent to the client, and the target search term is added to the target keyword account or not according to the adding feedback instruction of the client.

[0172] That is, when the target search word meets the preset adding condition, the target search word is not directly added to the account, but the keyword information matched by the target search word is used to generate adding information, which includes the recommended bid, recommended matching method, recommended account plan unit to be added, etc.

[0173] After receiving the word-adding information, the client of the advertiser adds or does not add the target search word to the account plan unit of the target keyword according to the word-adding instruction or the word-not-adding instruction fed back by the advertiser.

[0174] Step S400: When the target search term meets the preset negative word condition, the target search term is added to the account negative word library of the target keyword;

[0175] The negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword.

[0176] In an embodiment of the present invention, when a keyword with a high bid matches a keyword with a low bid, it results in that in a search term report, the average click price of the same search term as the keyword with the low bid exceeds its bid.

[0177] Assume that in the cpc delivery mode:

[0178] The bid for keyword I is 5 yuan, and the bid for keyword J is 2 yuan. When keyword I matches keyword J, in the search term report, the average click price of search term J, which is the same as keyword J, is 3 yuan, which actually exceeds the bid of 2 yuan preset by the advertiser for keyword J. Therefore, in the account plan unit of keyword I, it is best not to match keyword J again.

[0179] In addition, the negative word condition can also be set as a premium ratio. When the premium of search term J exceeds that of keyword J by a ratio higher than the premium ratio, the target search term is added to the negative word library of the target keyword's account.

[0180] Assuming that in an embodiment of the present invention, the premium ratio is 30%, the premium of search term J is 3-2=1 yuan, and the premium ratio is 1 / 2=50%. If the premium ratio exceeds 30%, search term J will be added to the negative word library in the planning unit where keyword I is located, so that keyword I will no longer match search term J in the future.

[0181] Assume that the advertiser sets the following high-to-low range standard:

[0182] The average CPC is more than 30% higher than the bid price;

[0183] The average click price is 3 yuan, which is 50% higher than the bid of 2 yuan. 50%>30%, which meets the standard of high-matching-low.

[0184] Therefore, keyword b is extracted separately for advertisers to review and judge whether to add it as a negative word in the management structure of keyword a.

[0185] Step S500: when the target search term meets the preset negative word condition, generate negative word information and send it to the client, and add the target search term to the negative word library of the target keyword account or not according to the negative word feedback instruction of the client;

[0186] In order to make negative words controllable and to be able to decide whether to add them to the negative word library based on the needs of advertisers, in an embodiment of the present invention, this step is added, that is, when a target search word meets a preset negative word condition, the target search word is not directly added to the negative word library, but the keyword information matched by the target search word is used to generate negative word information, and the negative word information includes a recommended bid, a recommended matching method, and a recommended account plan unit to be added.

[0187] After receiving the negative word information, the client of the advertiser adds or does not add the target search word to the negative word library of the account unit plan of the target keyword according to the negative word adding instruction or the negative word not adding instruction fed back by the advertiser.

[0188] Example 3

[0189] The embodiment of the present invention provides a system for automatically adding or removing words from bidding advertisements, as shown in the attached Figure 5 As shown, the system includes a processor, a memory, an interface device, a communication device, a display device, an input device, a speaker, a microphone, and the like.

[0190] The processor may be, for example, a central processing unit (CPU), a microprocessor (MCU), etc. The memory may include, for example, a ROM (read-only memory), a RAM (random access memory), a non-volatile memory such as a hard disk, etc. The interface device may include, for example, a USB interface, a headphone interface, etc. The communication device may be capable of wireless communication or wired communication. The display device may be, for example, a liquid crystal display, a touch display, etc. The input device may include, for example, a touch screen, a keyboard, etc. The user may output / input voice information through a speaker and a microphone.

[0191] The memory is used to store basic data and computer programs, and when the computer programs are executed by the processor, the aforementioned method of automatically adding or removing words from advertisements is implemented.

[0192] The system of the embodiment of the present invention includes:

[0193] A target search term acquisition unit, configured to acquire a target search term from at least one advertising platform;

[0194] A word adding judgment unit, used for adding the target search word to the account unit of the target keyword when the target search word meets the preset word adding condition;

[0195] The addition condition at least includes that the similarity between the target search term and the target keyword is greater than the addition threshold; and

[0196] A negative word judgment unit, used for adding the target search word to the account negative word library of the target keyword when the target search word meets the preset negative word condition;

[0197] The negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword.

[0198] A matching unit, used to convert a target keyword into a target element word according to a matching rule; and to convert a search word into an account element word according to a matching rule;

[0199] The matching rules include:

[0200] Establish root unit table and element unit table;

[0201] The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap;

[0202] The element unit table includes N element modules, which include numbers, letters and / or symbols, and different element modules have different contents. The element modules correspond to the root modules one by one and do not overlap.

[0203] Segment the target keywords and / or search terms in the account to obtain word roots;

[0204] The root word is converted into a target element word and / or an account element word according to the one-to-one correspondence between the root word unit table and the element unit table.

[0205] In an embodiment of the present invention, the word adding condition at least includes that the similarity between the target element word and the account element word is greater than the word adding threshold;

[0206] The calculation method of similarity a is:

[0207] a=x / y

[0208] Where x is the number of elements in the account element word that are identical to the target element word, and y is the number of elements in the target element word.

[0209] Those skilled in the art should understand that although multiple devices are shown in the figure, the computer device of the embodiment of the present invention may only involve some of the devices, or may also cover other devices. The technicians can design instructions according to the scheme disclosed by the present invention. How the instructions control the processor to operate is a well-known technology in the art, so it will not be described in detail here.

[0210] The present invention may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present invention.

[0211] It is well known to those skilled in the art that with the development of electronic information technology such as integrated circuit technology and the trend of software hardware, it has become more difficult to clearly divide the boundaries between system software and hardware. Because any operation can be implemented by software or by hardware. The execution of any instruction can be completed by hardware or by software. Whether a hardware implementation solution or a software implementation solution is adopted for a certain machine function depends on non-technical factors such as price, speed, reliability, storage capacity, and change cycle. Therefore, for ordinary technicians in the field of electronic information technology, a more direct and clear way to describe a technical solution is to describe the various operations in the solution. Knowing the operations to be performed, technicians in this field can directly design the desired product or device based on consideration of the non-technical factors.

[0212] Although some specific embodiments of the present invention have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present invention. It will be appreciated by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.

Claims

1. A method for automatically adding or removing words from bidding advertisements, characterized in that: include: Obtain target search terms through at least one advertising platform; Compare the similarity between the target search term and the keywords in the account, and send the comparison result to the client; When the target search term meets the preset adding condition, the target search term is added to the target keyword account; The word addition condition at least includes that the similarity between the target search word and the target keyword is greater than the word addition threshold; Convert the target search term into the target element term according to the matching rules; Converting keywords into account element words according to the matching rules; The word adding condition at least includes that the similarity between the target element word and the account element word is greater than the word adding threshold; The calculation method of the similarity a is: a=x / y Where x is the number of elements in the account element word that are identical to the target element word, and y is the number of elements in the target element word; Acquire the matched account element word according to the word addition condition; When the matched account element word is one, adding the target search word to the target keyword account corresponding to the account element word; When there are more than one account element words matched, the similarities of the account element words matched are compared, the one with the highest similarity is selected as the target keyword, and the target search word is added to the target keyword account; When there are more than one account element words with the highest similarity, the advertisement data of the account element words are compared, the highest weight in the advertisement data is selected as the target keyword, and the target search word is added to the target keyword account.

2. The method for automatically adding or removing words from bidding advertisements according to claim 1, characterized in that: The matching rules include: Establish root unit table and element unit table; The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap; The element unit table includes N element modules, the element modules include numbers, letters and / or symbols, and different element modules have different contents. The element modules correspond to the root modules one by one and do not overlap. Segment the target search term and / or the keyword in the account to obtain word roots; The root word is converted into a target element word and / or an account element word according to a one-to-one correspondence between the root word unit table and the element unit table.

3. The method for automatically adding or removing words from bidding advertisements according to any one of claims 1 to 2, characterized in that: Also includes: When the target search word meets the preset negative word condition, the target search word is added to the account negative word library of the target keyword; The negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword.

4. The method for automatically adding or removing words from bidding advertisements according to claim 2, characterized in that: Also includes: When at least one of the word roots obtained by the segmentation of the target search term cannot be converted into an element, negative word information is generated and sent to the client, and the word root that cannot be converted into an element is added or not added to the account negative word library according to the negative word feedback instruction of the client.

5. The method for automatically adding or removing words from bidding advertisements according to claim 4, characterized in that: The advertising data includes consumption, clicks, impressions, conversions, orders, inquiries, payments and payment amounts, and the weights are consumption, clicks, impressions, conversions, orders, inquiries, payments or payment amounts.

6. The method for automatically adding or removing words from bidding advertisements according to claim 5, characterized in that: Also includes: When the target search term meets the preset addition condition, the addition information is generated and sent to the client, and the target search term is added to the target keyword account or not according to the addition feedback instruction of the client; The word-added feedback instruction includes a word-added instruction or a no-word-added instruction.

7. The method for automatically adding or removing words from bidding advertisements according to claim 3, characterized in that: Also includes: When the target search term meets the preset negative word condition, negative word information is generated and sent to the client, and the target search term is added to the negative word library of the target keyword account or not added according to the negative word feedback instruction of the client; The negative word feedback instruction includes a negative word adding instruction and a negative word not adding instruction.

8. The method for automatically adding or removing words from bidding advertisements according to claim 1, characterized in that: The comparing the similarity between the target search term and the keyword in the account and sending the comparison result to the client comprises: Compare the similarity between the target search term and the keywords in the account, take the keyword with the highest similarity as the target keyword, generate word addition information and send it to the client, and add the target search term to the account of the target keyword or not according to the word addition feedback instruction of the client; The word-added feedback instruction includes a word-added instruction or a no-word-added instruction.

9. The method for automatically adding or removing words from bidding advertisements according to claim 8, characterized in that: When there is only one keyword with the highest similarity, the keyword with the highest similarity is used as the target keyword, and the added word information is generated and sent to the client; When there are more than one keywords with the highest similarity, the advertisement data of the keywords are compared, the keyword with the highest weight in the advertisement data is selected as the target keyword, and the added word information is generated and sent to the client.

10. The method for automatically adding or removing words from bidding advertisements according to claim 9, characterized in that: The advertising data includes consumption, clicks, impressions, conversions, orders, inquiries, payments and payment amounts, and the weights are consumption, clicks, impressions, conversions, orders, inquiries, payments or payment amounts.

11. A bidding advertisement automatic word addition and word rejection system, characterized in that: include: A target search term acquisition unit, configured to acquire a target search term from at least one advertising platform; A word adding judgment unit, configured to add the target search word to the target keyword account when the target search word meets a preset word adding condition; The word addition condition at least includes that the similarity between the target search word and the target keyword is greater than the word addition threshold; as well as a negative word judgment unit, configured to add the target search word to the account negative word library of the target keyword when the target search word meets a preset negative word condition; The negative word condition at least includes that the bid of the target search word is lower than the bid of the target keyword; A matching unit, used for converting a target search word into a target element word according to a matching rule; and / or Converting keywords into account element words according to the matching rules; The matching rules include: Establish root unit table and element unit table; The root unit table includes N root modules, each of which includes roots with the same or similar meanings, and roots in different root modules do not overlap; The element unit table includes N element modules, the element modules include numbers, letters and / or symbols, and different element modules have different contents. The element modules correspond to the root modules one by one and do not overlap. Segment the target search term and / or the keyword in the account to obtain word roots; The root word is converted into a target element word and / or an account element word according to a one-to-one correspondence between the root word unit table and the element unit table.

12. The bidding advertisement automatic word addition and word rejection system according to claim 11, characterized in that: The word adding condition at least includes that the similarity between the target element word and the account element word is greater than the word adding threshold; The calculation method of the similarity a is: a=x / y Where x is the number of elements in the account element word that are identical to the target element word, and y is the number of elements in the target element word.

Images