A Search Engine Evaluation Method and Device

Abstract

The present application discloses a search engine evaluation method and apparatus. The method includes: obtaining search terms; inputting each search term into a search engine to be evaluated to obtain a list of recommended terms corresponding to each search term output by the search engine; scoring the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term; and obtaining a comprehensive score of the search engine according to the scoring result. It can be seen that the present application can obtain recommended terms through search terms and a search engine, obtain a scoring result corresponding to each search term according to the recommended terms, and then obtain a scoring result of the search engine. Therefore, the present application can score the search engine according to the recommended terms, thereby completing the evaluation of the search engine.

Description

Technical Field

[0001] This application relates to the field of search engines, and more particularly, to a search engine evaluation method and apparatus. Background Art

[0002] Modern society is a society with convenient shopping, so that e-commerce is becoming a global economic entity, and most people shop through various online platforms or shopping apps.

[0003] Among them, users can search for the required products through the search engines of online platforms or shopping apps. Therefore, online platforms and various shopping apps cannot do without search engines. Online platforms and various shopping apps pay great attention to the performance of search engines. The performance of a search engine can be obtained through the evaluation results of the search engine. Therefore, how to evaluate search engines has become a research direction in the e-commerce field.

[0004] In summary, there is an urgent need to provide a search engine evaluation method that can help the e-commerce field evaluate search engines. Summary of the Invention

[0005] In view of this, this application provides a search engine evaluation method and apparatus for helping the e-commerce field evaluate search engines.

[0006] To achieve the above object, the following solutions are proposed:

[0007] A search engine evaluation method includes:

[0008] Obtain search terms;

[0009] Input each search term into the search engine to be evaluated, and obtain a list of recommended terms corresponding to each search term output by the search engine;

[0010] Score the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term;

[0011] Obtain the comprehensive score of the search engine according to the scoring result.

[0012] Optionally, the obtaining of the search terms includes:

[0013] Obtain a first set of search terms from a database that stores historical search terms input by users into the search engine;

[0014] Filter and clean the first set of search terms according to a preset rule to obtain a second set of search terms after filtering and cleaning;

[0015] Tokenize each search term in the second search term set, and the tokenization result is used as the final search term.

[0016] Optionally, the tokenizing each search term in the second search term set, and the tokenization result is used as the final search term, includes:

[0017] Tokenize each search term in the second search term set to obtain a tokenization result;

[0018] According to a preset rule, perform a priority ranking on the tokenization result to obtain a ranked tokenization result;

[0019] Select the tokenization results that meet the preset conditions from the ranked tokenization results as the search terms.

[0020] Optionally, the scoring the recommended word list corresponding to each search term to obtain a scoring result corresponding to each search term, includes:

[0021] According to the sorting order, click-through rate and / or conversion rate of each recommended word in the recommended word list corresponding to each search term, determine the score of the recommended word list as the scoring result corresponding to the corresponding search term.

[0022] Optionally, the scoring the recommended word list corresponding to each search term to obtain a scoring result corresponding to each search term, includes:

[0023] According to the inclusion situation of each recommended word in the recommended word list corresponding to each search term for the set type of content, determine the content score of each recommended word;

[0024] Comprehensively determine the scoring result of the search term corresponding to the recommended word list based on the content scores of each recommended word in the recommended word list.

[0025] Optionally, the scoring the recommended word list corresponding to each search term to obtain a scoring result corresponding to each search term, includes:

[0026] Input each recommended word into the search engine to obtain the search result corresponding to each recommended word;

[0027] According to the search result, determine the score of the recommended word corresponding to the search result;

[0028] Comprehensively determine the scoring result of the search term corresponding to the recommended word list based on the scores of each recommended word in the recommended word list.

[0029] Optionally, the determining the score of the recommended word corresponding to the search result according to the search result, includes:

[0030] Determine the number of search results corresponding to each recommended word;

[0031] Score the recommended terms according to the number of the search results to obtain the scores of the recommended terms corresponding to the search results.

[0032] Optionally, determining the scores of the recommended terms corresponding to the search results according to the search results includes:

[0033] Count the number of brands and categories in the search results of each recommended term to obtain the total count;

[0034] Score the recommended terms according to the total count to obtain the scores of the recommended terms corresponding to the search results.

[0035] Optionally, scoring the list of recommended terms corresponding to each search term to obtain the scoring result corresponding to each search term includes:

[0036] Determine the score of the list of recommended terms according to the sorting order of each recommended term in the list of recommended terms corresponding to each search term, as well as the click-through rate and conversion rate of each recommended term, and use it as the first score of the search term corresponding to the list of recommended terms;

[0037] Determine the content score of each recommended term according to the inclusion of each recommended term in the list of recommended terms in the set type of content;

[0038] Calculate the score of the list of recommended terms by comprehensively considering the content scores of each recommended term in the list of recommended terms, and use it as the second score of the search term corresponding to the list of recommended terms;

[0039] Input each recommended term into the search engine to obtain the search results corresponding to each recommended term;

[0040] Determine the number of the search results corresponding to each recommended term;

[0041] Score the recommended terms according to the number to determine the number score of the recommended terms;

[0042] Calculate the score of the list of recommended terms by comprehensively considering the number scores of each recommended term in the list of recommended terms, and use it as the third score of the search term corresponding to the list of recommended terms;

[0043] Count the number of brands and categories in the search results of each recommended term to obtain the total count;

[0044] Calculate the statistical score of the recommended terms according to the total count;

[0045] Calculate the score of the list of recommended terms by comprehensively considering the statistical scores of each recommended term in the list of recommended terms, and use it as the fourth score of the search term corresponding to the list of recommended terms;

[0046] Based on the first score, the second score, the third score, and the fourth score described above, the total score of the search term is calculated as the scoring result corresponding to the search term.

[0047] A search engine evaluation device includes:

[0048] A search term acquisition unit for acquiring a search term;

[0049] A recommended word list acquisition unit for inputting each search term into the search engine to be evaluated, and obtaining the recommended word list corresponding to each search term output by the search engine;

[0050] A first scoring unit for scoring the recommended word list corresponding to each search term to obtain the scoring result corresponding to each search term;

[0051] A total score acquisition unit for obtaining the comprehensive score of the search engine according to the scoring result.

[0052] From the above technical solutions, it can be seen that in this application, the search term can be acquired first, and then for each search term, the corresponding recommended word can be obtained according to the search engine to be evaluated, and each recommended word corresponding to each search term is scored to obtain the scoring result corresponding to each search term. Finally, based on this scoring result, the evaluation of the search engine is completed to obtain the comprehensive score of the search engine. It can be seen that this application can obtain the recommended word through the search term and the search engine, obtain the scoring result corresponding to each search term according to the recommended word, and then obtain the scoring result of the search engine. Therefore, this application can score the search engine according to the recommended word, thereby completing the evaluation of the search engine.

[0053] In addition, in this application, traffic evaluation is not required, and it is not necessary to evaluate the search engine through AB testing with conversion rate and click-through rate as indicators, reducing the influence of objective factors, and enabling the evaluation of the search engine to be evaluated before e-commerce users use it. Description of the Drawings

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only the embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0055] Figure 1 It is a flowchart of a search engine evaluation method disclosed in an embodiment of the present application;

[0056] Figure 2 It is a schematic diagram of a search engine exemplified in an embodiment of the present application;

[0057] Figure 3 This is a schematic structural diagram of a search engine evaluation device disclosed in an embodiment of the present application. Detailed implementation manners

[0058] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0059] Next, it will be combined with Figure 1 A detailed introduction to the search engine evaluation method of the present application will be given, including the following steps:

[0060] Step S110: Obtain search terms.

[0061] Specifically, there are multiple ways in the present application to obtain search terms. Here, one optional way is provided, that is, a search term database can be established in advance, and the words that can be used as search terms are stored in the search term database. When a search term is needed, a search term that meets the requirements is selected from the search term database.

[0062] Step S120: Input each search term into the search engine to be evaluated, and obtain a list of recommended words corresponding to each search term output by the search engine.

[0063] Specifically, as Figure 2 shown, this can be a schematic diagram of a search engine exemplified in an embodiment of the present application. The search term can be "lian", and when "lian" is input into the search engine, a list of recommended words corresponding to "lian" can be obtained.

[0064] Optionally, as Figure 2 shown, the words included in the list of recommended words can be: "dress", "romper", "red dress", "commuter dress", "short dress".

[0065] Optionally, the obtained search terms can include multiple words. One possibility provided by the present application is that there are a total of 3 words in the obtained search terms.

[0066] Then, at this time, the first search term can be input into the search engine, and a list of recommended words corresponding to the first search term output by the search engine is obtained, and the list of recommended words is stored in the memory for subsequent steps to evaluate the list of recommended words.

[0067] At this time, the second search term can be input into the search engine to obtain a list of recommended terms corresponding to the second search term output by the search engine, and this list of recommended terms is stored in the memory for subsequent steps to evaluate this list of recommended terms.

[0068] At this time, the third search term can be input into the search engine to obtain a list of recommended terms corresponding to the third search term output by the search engine, and this list of recommended terms is stored in the memory for subsequent steps to evaluate this list of recommended terms.

[0069] Step S130: Score the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term.

[0070] Specifically, the obtained search terms may contain multiple words, so the corresponding list of search terms also has multiple. One possibility provided by this application is that there are a total of 3 words in the obtained search terms, then there are three lists of search terms, and there is a one-to-one correspondence between the search terms and the lists of search terms.

[0071] At this time, the three lists of search terms can be scored respectively.

[0072] Moreover, since there is a one-to-one correspondence between the list of search terms and the search terms, the scoring result of the list of search terms and the search terms also has a one-to-one correspondence.

[0073] Step S140: Obtain the comprehensive score of the search engine according to the scoring result.

[0074] Specifically, the obtained search terms may contain multiple words, so the corresponding list of search terms also has multiple, and the obtained scoring results also have multiple. One possibility provided by this application is that there are a total of 3 words in the obtained search terms, then there are also three scoring results, and there is a one-to-one correspondence between the search terms and the frequency division results.

[0075] At this time, the comprehensive score of the search engine can be obtained according to these three scoring results corresponding to the search terms.

[0076] Among them, there are various ways to obtain the comprehensive score of the search engine according to the scoring result. For example, the individual scoring results can be directly added, or the individual scoring results can be added according to their weights according to the weights of the individual scoring results.

[0077] As can be seen from the above technical solutions, the search engine evaluation method provided by the embodiments of the present application first obtains search terms, then obtains the recommended terms corresponding to each search term according to the search engine to be evaluated, scores the recommended terms corresponding to each search term, obtains the scoring results corresponding to each search term, and finally, based on the scoring results, completes the evaluation of the search engine to obtain the comprehensive score of the search engine. It can be seen that the present application obtains recommended terms through search terms and the search engine, obtains the score results corresponding to each search term according to the recommended terms, and then obtains the score result of the search engine. Therefore, the present application scores the search engine according to the recommended terms, thereby completing the evaluation of the search engine.

[0078] In addition, in the present application, it is not necessary to perform traffic evaluation, nor to evaluate the search engine by using conversion rate and click-through rate as indicators through AB tests, reducing the influence of objective factors, and being able to evaluate the search engine to be evaluated before e-commerce users use it.

[0079] Further, in some embodiments of the present application, the process of step S110, obtaining search terms, is described in detail as follows:

[0080] S10. Obtain a first set of search terms from a database, where the database stores historical search terms input by users into the search engine.

[0081] Specifically, a database can be established in advance, and the database stores historical search terms input by users in the search engine.

[0082] When it is necessary to collect search terms, the set of search terms can be obtained from this database.

[0083] S11. Filter and clean the first set of search terms according to preset rules to obtain a second set of search terms after filtering and cleaning.

[0084] Specifically, the preset rules can filter and clean words that make it difficult for the search engine to provide recommended terms, such as invalid words, general words, overly long words, etc.

[0085] Among them, invalid words can refer to words composed only of numbers or only of symbols. General words can refer to words without a target orientation, such as "male" and "female". Overly long words can refer to words whose number of characters exceeds a preset threshold.

[0086] S12. Segment each search term in the second set of search terms, and use the segmentation result as the final search term.

[0087] Specifically, segmentation can include prefix extraction, word segmentation extraction, etc.

[0088] The prefix extraction may extract the first character of each search term in the second search term set. For example, the result of prefix extraction of "掌" is "连".

[0089] However, the word segmentation extraction can split the constituent words of each search word in the second search word set. For example, the result of the word segmentation extraction of "red dress" is "red" and "dress".

[0090] It can be found from the above technical solution that this embodiment adds the process of filtering and cleaning the search words and performing word segmentation compared to the above embodiment. In this way, more practical search words that can better test the performance of the search engine can be obtained, and the search engine can be better evaluated.

[0091] Furthermore, considering the problem that there may be too many words after word segmentation, it is necessary to select some of the word segmentation results as search words. Specifically, in some embodiments of the present application, step S12, the process of performing word segmentation on each search word in the second search word set and using the word segmentation results as the final search word, is described in detail, and the steps are as follows:

[0092] S120: Segment each search word in the second search word set to obtain a segmentation result.

[0093] Specifically, word segmentation may include prefix extraction, word segmentation extraction, etc., to obtain prefix extraction results and word segmentation extraction results of each search word in the second search word set.

[0094] S121. Prioritize the word segmentation results according to preset rules to obtain prioritized word segmentation results.

[0095] Specifically, the preset rules can have multiple sorting methods, such as using TF / IDF and Kmeans algorithms for sorting, or sorting the word segmentation results according to the number of user inputs, the number of purchases and favorites of the products corresponding to the word segmentation results. On this basis, the word segmentation results can also be sorted according to seasons and regions.

[0096] S122. Select a word segmentation result that meets a preset condition from the sorted word segmentation results as a search term.

[0097] Specifically, there may be multiple situations for the preset conditions. For example, the first 100 sorted segmentation results may be used as search terms, or the first 50 sorted segmentation results and each search term in the second search term set before segmentation may be formed into a set as the search term.

[0098] As can be seen from the above technical solutions, in this embodiment, a process of sorting and selecting the word segmentation results according to preset rules is added compared with the above embodiment, which can select search terms that can better test the performance of the search engine and can better evaluate the search engine.

[0099] In some embodiments of the present application, the process of step S130 of scoring the recommended word list corresponding to each search term to obtain the scoring result corresponding to each search term is described in detail as follows:

[0100] S21. Determine the score of the recommended word list as the scoring result corresponding to the search term according to the sorting order, click-through rate and / or conversion rate of each recommended word in the recommended word list corresponding to each search term.

[0101] Specifically, when the search term is input into the search engine, the recommended word list corresponding to the search term can be obtained, as Figure 2 shown.

[0102] Among them, the recommended words in the recommended word list can have a sorting order, and the closer to the search engine input box, the higher the sorting. For example, as Figure 2 shown, the sorting order of the recommended word list is "dress", "romper", "red dress", "commuting dress", "short dress".

[0103] In fact, each recommended word has its corresponding click-through rate and conversion rate.

[0104] Among them, the click-through rate can refer to the number of times the user clicks on the recommended word, and the conversion rate can refer to the number of times the user purchases the product corresponding to the recommended word.

[0105] In this way, the score of the recommended word list can be determined according to the sorting order, click-through rate and / or conversion rate of each recommended word in the recommended word list corresponding to each search term.

[0106] For example, sort the recommended words in the recommended word list according to the click-through rate and / or conversion rate of each recommended word, and match the sorting result with the sorting order in the recommended word list, and score the recommended word list according to the preset sorting scoring rules.

[0107] Among them, there can be various situations for the preset sorting scoring rules. The present application provides an optional method. For example, it can be preset that the full score of the recommended word list is 10 points, and the 10 points are evenly distributed to each recommended word in the recommended word list. If the order of the recommended word in the sorting result matches the sorting order in the recommended word list, the score of the recommended word can be obtained.

[0108] As can be seen from the above technical solution, in this embodiment, a process of scoring the recommendation word list is added compared with the above embodiment according to the sorting order of each recommendation word in the recommendation word list, the click-through rate of each recommendation word, and / or the conversion rate of each recommendation word. It is possible to score the recommendation word list from the sorting order of each recommendation word in the recommendation word list, so as to obtain the evaluation result of the search engine.

[0109] The above embodiment provides an optional implementation manner of step S130. Considering that the recommendation words in the recommendation word list may be composed of various types of content, and the richer the type of content included in the recommendation word, the easier it is to help users find suitable products, and the more in line with the requirements of the e-commerce field for the search engine. Therefore, the score of each recommendation word can be determined according to the composition of the type of content of the recommendation word. Based on this, the present application provides another optional implementation manner of step S130, and the specific steps are as follows:

[0110] S31. Determine the content score of each recommendation word according to the inclusion of each recommendation word in the recommendation word list corresponding to each search term in the set type of content.

[0111] Specifically, the set type of content may include product words such as dresses, shoes, etc., may also include attribute words such as red, short sleeves, etc., and may also include brand names.

[0112] Among them, the content score of each recommendation word can be determined according to the richness of the recommendation word.

[0113] Specifically, the richness may refer to the richness of the set type of content included. That is, the content score of a recommendation word that includes both product words and brand names and attribute words is higher than that of a recommendation word that only includes product words. If the type of content included in the recommendation word is richer and more in line with the requirements of the e-commerce field for the search engine, the score can also be higher.

[0114] S32. Synthesize the content scores of each recommendation word in the recommendation word list to determine the scoring result of the search term corresponding to the recommendation word list.

[0115] Specifically, there are various ways to determine the scoring result of the search term corresponding to the recommendation word list according to the content scores of each recommendation word.

[0116] For example, the content scores of each recommendation word in the recommendation word list can be directly added to obtain the scoring result corresponding to the recommendation word list, so as to obtain the scoring result of the search term corresponding to the recommendation word list.

[0117] Similarly, according to the weights of each recommendation word in the recommendation word list, the content scores of each recommendation word in the recommendation word list can be added to obtain the scoring result corresponding to the recommendation word list, so as to obtain the scoring result of the search term corresponding to the recommendation word list.

[0118] It can be found from the above technical solution that in this embodiment, the content score of each recommended word can be determined according to the inclusion of each recommended word in the recommended word list in the set type of content, so as to obtain the scoring result of the search term corresponding to the recommended word list.

[0119] The above embodiment provides an optional implementation manner of step S130. Considering that if each recommended word is input into the search engine, search results can be obtained, and the search results can help users find the required products, which meets the requirements of the e-commerce field for the search engine. Based on this, the present application provides another optional implementation manner of step S130, and the specific steps are as follows:

[0120] S41. Input each recommended word into the search engine to obtain the search result corresponding to each recommended word.

[0121] Specifically, input each recommended word into the search engine, search for the recommended word, and obtain the corresponding search result of the recommended word.

[0122] S42. Determine the score of the recommended word corresponding to the search result according to the search result.

[0123] Specifically, the recommended word can be scored according to the search result of the recommended word.

[0124] S43. Synthesize the scores of the recommended words in the recommended word list to determine the scoring result of the search term corresponding to the recommended word list.

[0125] Specifically, there are various ways to determine the scoring result of the search term corresponding to the recommended word list according to the scores of the recommended words.

[0126] For example, the scores of the recommended words in the recommended word list can be directly added to obtain the scoring result corresponding to the recommended word list, so as to obtain the scoring result of the search term corresponding to the recommended word list.

[0127] Similarly, according to the weights of the recommended words in the recommended word list, the scores of the recommended words in the recommended word list can be added to obtain the scoring result corresponding to the recommended word list, so as to obtain the scoring result of the search term corresponding to the recommended word list.

[0128] It can be found from the above technical solution that in this embodiment, the process of determining the score of each recommended word according to the search result of each recommended word in the recommended word list, so as to obtain the scoring result of the search term corresponding to the recommended word list, can score the recommended word list from the search results of the recommended words in the recommended word list, so as to obtain the evaluation result of the search engine.

[0129] In some embodiments of the present application, there are various implementation manners for the process of step S42, determining the scores of the recommended words corresponding to the search results. The present application provides two of these implementation manners, specifically as follows:

[0130] The first one:

[0131] S420. Determine the number of search results corresponding to each recommended word.

[0132] Specifically, it can be calculated how many products are included in the search results corresponding to each recommended word.

[0133] S421. Score the recommended words according to the number of the search results to obtain the scores of the recommended words corresponding to the search results.

[0134] Specifically, there are various ways to score the recommended words according to the number of the search results. For example, if the number exceeds a preset threshold, the recommended word can get a full score; otherwise, the recommended word gets no score.

[0135] Similarly, it can also be set that when the number is 100, the recommended word is full score of 100 points, then one search result corresponds to 1 point. Then, if the number of search results of the recommended word is 58, it is 58 points.

[0136] The second one:

[0137] S422. Count the number of brands and categories in the search results of each recommended word to obtain the total count.

[0138] Specifically, it can be calculated how many brands and categories are included in the search results corresponding to each recommended word to obtain the total count.

[0139] S423. Score the recommended words according to the total count to obtain the scores of the recommended words corresponding to the search results.

[0140] Specifically, there are various ways to score the recommended words according to the total count. For example, if the total count exceeds a preset threshold, the recommended word can get a full score; otherwise, the recommended word gets no score.

[0141] Similarly, it can also be set that when the total count is 50, the recommended word is full score of 100 points, then one search result corresponds to 2 points. Then, if the number of search results of the recommended word is 8, it is 16 points.

[0142] It can be found from the above technical solutions that this embodiment adds the process of determining the score of each recommended word according to the search results of the recommended words compared with the above embodiment. The scores of each recommended word can be determined according to the number of search results or according to the number of brand categories in the search results, so as to obtain the evaluation result of the search engine.

[0143] The above embodiments provide one optional implementation manner of step S130. Considering that several implementation manners of step S130 in the above embodiments can be combined to obtain a more objective and comprehensive evaluation result of the search engine. Based on this, the present application provides another optional implementation manner of step S130, and the specific steps are as follows:

[0144] S71. Determine the score of the recommended word list as the first score of the search term corresponding to the recommended word list according to the sorting order of each recommended word in the recommended word list corresponding to each search term, as well as the click-through rate and conversion rate of each recommended word. Corresponding to the foregoing step S21, for details, refer to the foregoing introduction and will not be elaborated here.

[0145] S72. Determine the content score of each recommended word according to the inclusion of each recommended word in the recommended word list in the set type of content. Corresponding to the foregoing step S31, for details, refer to the foregoing introduction and will not be elaborated here.

[0146] S73. Calculate the score of the recommended word list by comprehensively considering the content scores of each recommended word in the recommended word list as the second score of the search term corresponding to the recommended word list. Corresponding to the foregoing step S32, for details, refer to the foregoing introduction and will not be elaborated here.

[0147] S74. Input each recommended word into the search engine to obtain the search results corresponding to each recommended word. Corresponding to the foregoing step S41, for details, refer to the foregoing introduction and will not be elaborated here.

[0148] S75. Determine the number of search results corresponding to each recommended word. Corresponding to the foregoing step S420, for details, refer to the foregoing introduction and will not be elaborated here.

[0149] S76. Score the recommended words according to the number to determine the number score of the recommended words. Corresponding to the foregoing step S421, for details, refer to the foregoing introduction and will not be elaborated here.

[0150] S77. Calculate the score of the recommended word list by comprehensively considering the number scores of each recommended word in the recommended word list as the third score of the search term corresponding to the recommended word list.

[0151] Specifically, there are various ways to calculate the score of the recommended word list according to the number scores of each recommended word.

[0152] For example, the number scores of each recommended word in the recommended word list can be directly added to obtain the score corresponding to the recommended word list.

[0153] Similarly, based on the weights of each recommended term in the recommended term list, the scores of each recommended term in the recommended term list can be added with weights to obtain the scoring result corresponding to the recommended term list, thereby obtaining the scoring result of the search term corresponding to the recommended term list.

[0154] S78. Count the number of brands and categories in the search results of each recommended term to obtain the total count. Corresponding to the aforementioned step S422, refer to the aforementioned introduction in detail and will not be elaborated here.

[0155] S79. Calculate the statistical score of the recommended term based on the total count. Corresponding to the aforementioned step S423, refer to the aforementioned introduction in detail and will not be elaborated here.

[0156] S80. Synthesize the statistical scores of each recommended term in the recommended term list to calculate the score of the recommended term list, which is used as the fourth score of the search term corresponding to the recommended term list.

[0157] Specifically, there are various ways to calculate the score of the recommended term list based on the statistical scores of each recommended term.

[0158] For example, the statistical scores of each recommended term in the recommended term list can be directly added to obtain the score corresponding to the recommended term list.

[0159] Similarly, based on the weights of each recommended term in the recommended term list, the statistical scores of each recommended term in the recommended term list can be added with weights to obtain the score corresponding to the recommended term list.

[0160] S81. Synthesize the first score, the second score, the third score, and the fourth score to calculate the total score of the search term, which is used as the scoring result corresponding to the search term.

[0161] Specifically, there are various ways to calculate the total score of the search term based on the first score, the second score, the third score, and the fourth score.

[0162] For example, the first score, the second score, the third score, and the fourth score can be directly added to obtain the score corresponding to the search term.

[0163] Similarly, based on the weights of each score in the scoring result of the search term, the scores can be added with weights to obtain the scoring result corresponding to the search term.

[0164] As can be seen from the above technical solution, in this embodiment, the search terms are scored from multiple perspectives. The sorting order of the recommended word list, the composition of the recommended words, the number of search results of the recommended words, and the brand categories of the search results of the recommended words can all affect the final scoring result of the search engine, making the evaluation result of the recommended word list more objective, so as to better evaluate the search engine.

[0165] Next, the search engine evaluation device provided by the embodiments of the present application will be described. The search engine evaluation device described below can be correspondingly referred to the search engine evaluation method described above.

[0166] First, in combination with Figure 3 , the search engine evaluation device applied to the e-commerce field will be introduced. As Figure 3 shown, the information recommendation device may include:

[0167] A search term acquisition unit 100, configured to acquire search terms;

[0168] A recommended word list acquisition unit 110, configured to input each search term into the search engine to be evaluated, and obtain a recommended word list corresponding to each search term output by the search engine;

[0169] A scoring unit 120, configured to score the recommended word list corresponding to each search term, and obtain a scoring result corresponding to each search term;

[0170] A total score acquisition unit 130, configured to obtain a comprehensive score of the search engine according to the scoring result.

[0171] Optionally, the search term acquisition unit may include:

[0172] A first search term acquisition unit, configured to acquire a first search term set from a database, where the database stores historical search terms input by users into the search engine;

[0173] A filtering unit, configured to filter and clean the first search term set according to a preset rule, and obtain a second search term set after filtering and cleaning;

[0174] A word segmentation unit, configured to perform word segmentation on each search term in the second search term set, and the word segmentation result is used as the final search term.

[0175] Optionally, the word segmentation unit may include:

[0176] A second search term acquisition unit, configured to perform word segmentation on each search term in the second search term set, and obtain a word segmentation result;

[0177] A sorting unit, configured to perform a priority sorting on the word segmentation results according to a preset rule to obtain sorted word segmentation results;

[0178] A third search term obtaining unit, configured to select word segmentation results that meet a preset condition from the sorted word segmentation results as search terms.

[0179] Optionally, the scoring unit may include:

[0180] A first scoring unit, configured to determine a score of a recommended word list as a scoring result corresponding to a search term according to the sorting order, click-through rate, and / or conversion rate of each recommended word in the recommended word list corresponding to each search term.

[0181] Optionally, the scoring unit may also include:

[0182] A second scoring unit, configured to determine a content score of each recommended word according to the inclusion of each recommended word in the recommended word list corresponding to each search term in a set type of content;

[0183] A first comprehensive scoring unit, configured to comprehensively determine a scoring result of a search term corresponding to the recommended word list according to the content scores of each recommended word in the recommended word list.

[0184] Optionally, the scoring unit may further include:

[0185] A first search result obtaining unit, configured to input each recommended word into the search engine to obtain a search result corresponding to each recommended word;

[0186] A scoring unit, configured to determine a score of a recommended word corresponding to the search result according to the search result;

[0187] A second comprehensive scoring unit, configured to comprehensively determine a scoring result of a search term corresponding to the recommended word list according to the scores of each recommended word in the recommended word list.

[0188] Optionally, the scoring unit may include:

[0189] A first quantity determining unit, configured to determine the quantity of search results corresponding to each recommended word;

[0190] A first scoring unit, configured to score a recommended word according to the quantity of the search result to obtain a score of the recommended word corresponding to the search result.

[0191] Optionally, the scoring unit may further include:

[0192] A first total number statistical unit, configured to count the quantity of brands and categories in the search results of each recommended word to obtain a statistical total;

[0193] A second scoring unit, configured to score the recommended words according to the total count, and obtain the scores of the recommended words corresponding to the search results.

[0194] Optionally, the scoring unit may further include:

[0195] A fourth scoring unit, configured to determine the score of the recommended word list according to the sorting order of each recommended word in the recommended word list corresponding to each search term, as well as the click-through rate and conversion rate of each recommended word, and use it as the first score of the search term corresponding to the recommended word list;

[0196] A content score determination unit, configured to determine the content score of each recommended word according to the inclusion of each recommended word in the recommended word list in the set type of content;

[0197] A third integration unit, configured to integrate the content scores of each recommended word in the recommended word list, calculate the score of the recommended word list, and use it as the second score of the search term corresponding to the recommended word list;

[0198] A second search result acquisition unit, configured to input each recommended word into the search engine to obtain the search results corresponding to each recommended word;

[0199] A second quantity determination unit, configured to determine the quantity of the search results corresponding to each recommended word;

[0200] A quantity score determination unit, configured to score the recommended words according to the quantity, and determine the quantity scores of the recommended words;

[0201] A fourth integration unit, configured to integrate the quantity scores of each recommended word in the recommended word list, calculate the score of the recommended word list, and use it as the third score of the search term corresponding to the recommended word list;

[0202] A second total count unit, configured to count the quantity of brands and categories in the search results of each recommended word to obtain a total count;

[0203] A statistical score calculation unit, configured to calculate the statistical score of the recommended word according to the total count;

[0204] A fifth integration unit, configured to integrate the statistical scores of each recommended word in the recommended word list, calculate the score of the recommended word list, and use it as the fourth score of the search term corresponding to the recommended word list;

[0205] A sixth integration unit, configured to integrate the first score, the second score, the third score, and the fourth score, calculate the total score of the search term, and use it as the scoring result corresponding to the search term.

[0206] Optionally, the refined functions and extended functions of the program may refer to the above description.

[0207] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the said element.

[0208] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other.

[0209] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. The various embodiments of the present application can be combined with each other. Therefore, the present application will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for evaluating a search engine, characterized in that, it includes: Obtain search terms; Input each search term into the search engine to be evaluated, and obtain a list of recommended terms corresponding to each search term output by the search engine; Score the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term; According to the scoring results, obtain the comprehensive score of the search engine; The scoring the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term includes: Determine the content score of each recommended term according to the inclusion of each recommended term in the list of recommended terms corresponding to each search term in the set type of content; Comprehensively determine the scoring result of the search term corresponding to the list of recommended terms based on the content scores of the recommended terms in the list of recommended terms.

2. The method according to claim 1, characterized in that, the obtaining the search terms includes: Obtain a first set of search terms from a database that stores historical search terms input by users into the search engine; Filter and clean the first set of search terms according to preset rules to obtain a second set of search terms after filtering and cleaning; Perform word segmentation on each search term in the second set of search terms, and the word segmentation result is used as the final search term.

3. The method according to claim 2, characterized in that, the performing word segmentation on each search term in the second set of search terms, and the word segmentation result is used as the final search term, includes: Perform word segmentation on each search term in the second set of search terms to obtain a word segmentation result; Sort the word segmentation results according to preset rules to obtain a sorted word segmentation result; Select the word segmentation results that meet the preset conditions from the sorted word segmentation results as search terms.

4. The method according to claim 1, characterized in that, the scoring the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term includes: Determine the score of the list of recommended terms according to the sorting order, click-through rate and / or conversion rate of each recommended term in the list of recommended terms corresponding to each search term, and use it as the scoring result of the corresponding search term.

5. The method according to claim 1, characterized in that, the scoring the list of recommended terms corresponding to each search term to obtain a scoring result corresponding to each search term includes: Input each recommended term into the search engine to obtain the search results corresponding to each recommended term; Determine the score of the recommended term corresponding to the search results according to the search results; Comprehensively determine the scoring result of the search term corresponding to the list of recommended terms based on the scores of the recommended terms in the list of recommended terms.

6. The method according to claim 5, characterized in that, the determining the score of the recommended term corresponding to the search results according to the search results includes: Determine the number of search results corresponding to each recommended term; Score the recommended terms according to the number of search results to obtain the score of the recommended term corresponding to the search results.

7. The method according to claim 5, characterized in that, Determining the scores of the recommended words corresponding to the search results includes: Counting the number of brands and categories in the search results of each recommended word to obtain a total count; Rating the recommended words according to the total count to obtain the scores of the recommended words corresponding to the search results.

8. The method according to claim 1, wherein, Rating the list of recommended words corresponding to each search word to obtain the rating result corresponding to each search word includes: Determining the score of the recommended word list based on the sorting order of each recommended word, the click-through rate and conversion rate of each recommended word in the recommended word list corresponding to each search word, and using it as the first score of the search word corresponding to the recommended word list; Determining the content score of each recommended word according to the inclusion of each recommended word in the recommended word list in the content of the set type; Calculating the score of the recommended word list by comprehensively considering the content scores of each recommended word in the recommended word list, and using it as the second score of the search word corresponding to the recommended word list; Inputting each recommended word into the search engine to obtain the search results corresponding to each recommended word; Determining the number of search results corresponding to each recommended word; Rating the recommended words according to the number to determine the number score of the recommended words; Calculating the score of the recommended word list by comprehensively considering the number scores of each recommended word in the recommended word list, and using it as the third score of the search word corresponding to the recommended word list; Counting the number of brands and categories in the search results of each recommended word to obtain a total count; Calculating the statistical score of the recommended word according to the total count; Calculating the score of the recommended word list by comprehensively considering the statistical scores of each recommended word in the recommended word list, and using it as the fourth score of the search word corresponding to the recommended word list; Calculating the total score of the search word by comprehensively considering the first score, the second score, the third score and the fourth score, and using it as the rating result corresponding to the search word.

9. A search engine evaluation device, wherein, it includes: A search word acquisition unit for acquiring search words; A recommended word list acquisition unit for inputting each search word into the search engine to be evaluated to obtain the list of recommended words corresponding to each search word output by the search engine; A first rating unit for rating the list of recommended words corresponding to each search word to obtain the rating result corresponding to each search word; A total score acquisition unit for obtaining the comprehensive score of the search engine according to the rating result; wherein, the first rating unit includes: A second rating unit for determining the content score of each recommended word according to the inclusion of each recommended word in the recommended word list corresponding to each search word in the content of the set type; A first comprehensive score unit for comprehensively considering the content scores of each recommended word in the recommended word list to determine the rating result of the search word corresponding to the recommended word list.

Images