Methods, apparatus, computer equipment, and storage media for determining related terms

Abstract

This application discloses a method, apparatus, computer device, storage medium, and computer program product for determining related terms, relating to the field of big data intelligent analysis technology. The method includes: responding to a search request initiated by a business terminal, the search request carrying search keywords; obtaining business orders associated with the search keywords; determining at least one first related term associated with the search keywords based on business data values ​​in the business orders; obtaining a first item click-through rate (CTR) table matching the search keywords, and determining at least one second related term associated with the search keywords from the first item CTR table; and determining target related terms for the search keywords from at least one first related term and at least one second related term based on the business data values ​​of the first related terms and the CTR of the second related terms. This method can improve the accuracy of related terms.

Description

Technical Field

[0001] This application relates to the field of big data intelligent analysis technology, and in particular to a method, apparatus, computer device, storage medium and computer program product for determining related words. Background Technology

[0002] With the rapid development of e-commerce, more and more users and merchants are completing transactions through e-commerce platforms. When shopping online, users can enter keywords to search for items of interest through the search portal, or search based on keywords provided by the shopping platform.

[0003] However, currently, when users search for items, the system calculates the edit distance between the user's input words and historical keywords, and then provides the user with keywords that have a small edit distance to choose from, resulting in low accuracy of the determined related words. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer device, computer-readable storage medium, and computer program product for determining related words that can improve the accuracy of related words in response to the above-mentioned technical problems.

[0005] Firstly, this application provides a method for determining related terms. The method includes:

[0006] Responding to a search request initiated by a business terminal, the search request carrying search keywords;

[0007] Retrieve business orders associated with the search keywords;

[0008] Determine at least one first related term associated with the search keyword based on the business data values ​​in the business order;

[0009] Obtain a table of item click-through rates that match the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table;

[0010] Based on the business data value of the first related word and the click rate of the second related word, the target related word of the search keyword is determined from the at least one first related word and the at least one second related word.

[0011] In one embodiment, obtaining the click-through rate table of items matching the search keyword includes:

[0012] If there is no item click-through rate table matching the search keyword in the preset database, the search keyword is split to obtain at least two sub-keywords;

[0013] Based on the at least two sub-keywords, a matching item click-through rate table is determined from the preset database.

[0014] In one embodiment, determining at least one second related term associated with the search keyword from the item click-through rate table includes:

[0015] Determine the third associated word associated with each of the sub-keywords from the item click-through rate table;

[0016] Based on the click-through rate of each of the third related terms, a weighted average is applied to determine at least one second related term that is associated with the search keyword from among the third related terms.

[0017] In one embodiment, obtaining the business order associated with the search keyword includes:

[0018] If there is no business order in the business order database that is associated with the search keyword, then extract the characters of the search field from each business order in the business order database;

[0019] If the similarity between the characters in the search field and the search keyword reaches a preset value, and the pinyin of the characters in the search field is the same as that of the search keyword, then the business order corresponding to the search field will be the business order associated with the search keyword.

[0020] In one embodiment, determining the target related words for the search keyword from the at least one first related word and the at least one second related word based on the business data value of the first related word and the click-through rate of the second related word includes:

[0021] The association value of each associated word is obtained by weighting the business data value of the first associated word and the click rate of the second associated word.

[0022] The at least one first related word and the at least one second related word are sorted according to the association value, and the target related words of the search keyword are determined in descending order of association value.

[0023] In one embodiment, prior to obtaining the business order associated with the search keyword, the method further includes:

[0024] Obtain the format of the search keywords;

[0025] If the format of the search keyword does not conform to the preset format, the search keyword is converted to obtain the search keyword in the preset format.

[0026] Secondly, this application also provides a device for determining related terms. The device includes:

[0027] The response module is used to respond to search requests initiated by business terminals, wherein the search requests carry search keywords;

[0028] The order acquisition module is used to acquire business orders associated with the search keywords;

[0029] The first determining module is used to determine at least one first associated word associated with the search keyword based on the business data value in the business order;

[0030] The second determining module is used to obtain an item click-through rate table that matches the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table;

[0031] The third determining module is used to determine the target related words of the search keyword from the at least one first related word and the at least one second related word based on the business data value of the first related word and the click rate of the second related word.

[0032] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0033] Responding to a search request initiated by a business terminal, the search request carrying search keywords;

[0034] Retrieve business orders associated with the search keywords;

[0035] Determine at least one first related term associated with the search keyword based on the business data values ​​in the business order;

[0036] Obtain a table of item click-through rates that match the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table;

[0037] Based on the business data value of the first related word and the click rate of the second related word, the target related word of the search keyword is determined from the at least one first related word and the at least one second related word.

[0038] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0039] Responding to a search request initiated by a business terminal, the search request carrying search keywords;

[0040] Retrieve business orders associated with the search keywords;

[0041] Determine at least one first related term associated with the search keyword based on the business data values ​​in the business order;

[0042] Obtain a table of item click-through rates that match the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table;

[0043] Based on the business data value of the first related word and the click rate of the second related word, the target related word of the search keyword is determined from the at least one first related word and the at least one second related word.

[0044] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0045] Responding to a search request initiated by a business terminal, the search request carrying search keywords;

[0046] Retrieve business orders associated with the search keywords;

[0047] Determine at least one first related term associated with the search keyword based on the business data values ​​in the business order;

[0048] Obtain a table of item click-through rates that match the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table;

[0049] Based on the business data value of the first related word and the click rate of the second related word, the target related word of the search keyword is determined from the at least one first related word and the at least one second related word.

[0050] The aforementioned methods, apparatus, computer equipment, storage media, and computer program products for determining related terms, when determining related terms carrying search keywords, determine at least one first related term associated with the search keyword based on business orders associated with the search keyword, and at least one second related term associated with the search keyword from an item click-through rate (CTR) table. When determining related terms for search keywords based on CTR, the corresponding item CTR table is further matched based on the search keyword. This allows for the determination of corresponding item CTR tables for different search keywords, further accurately determining related terms associated with the search keyword from the item CTR dimension. This avoids the situation where matching is limited to specific search keywords, which may not meet the needs of actual scenarios and lead to inaccurate determination of the second related term. The target related term for the search keyword is determined based on the business data value of the first related term and the CTR of the second related term. When determining the target related term for the search keyword, and when determining related terms for the search keyword based on historical search keywords, the related terms for the search keyword are confirmed from both the dimensions of business data values ​​and item CTR, avoiding the determination of target related terms based on a single dimension and improving the accuracy of related terms. Attached Figure Description

[0051] Figure 1 This is a diagram illustrating the application environment of a method for determining related terms in one embodiment.

[0052] Figure 2 This is a flowchart illustrating a method for determining related terms in one embodiment;

[0053] Figure 3 This is a flowchart illustrating a method for determining the click-through rate of items based on search keywords in one embodiment.

[0054] Figure 4 This is a flowchart illustrating a method for determining related keywords based on click-through rate in one embodiment;

[0055] Figure 5 This is a flowchart illustrating a method for determining business orders associated with search keywords in another embodiment;

[0056] Figure 6 This is a flowchart illustrating the method for determining related terms in another embodiment;

[0057] Figure 7 This is a structural block diagram of a device for determining related words in one embodiment;

[0058] Figure 8 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0059] It is understood that the acquisition, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations.

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0061] The method for determining related terms provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, the business terminal 102 communicates with the server 104 via a network. A data storage system can store the data that the server 104 needs to process. The data storage system can be integrated onto the server 104 or placed on a cloud or other network server. The system responds to search requests initiated by the business terminal, the search request carrying search keywords; retrieves business orders associated with the search keywords; determines at least one first related word associated with the search keywords based on the business data values ​​in the business orders; retrieves a click-through rate table of items matching the search keywords, and determines at least one second related word associated with the search keywords from the item click-through rate table; and determines the target related word of the search keywords from at least one first related word and at least one second related word based on the business data values ​​of the first related word and the click-through rate of the second related word. The business terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0062] In one embodiment, such as Figure 2 As shown, a method for determining conjunctions is provided, which can be applied to... Figure 1 Taking the server in the example, the following steps are included:

[0063] Step 202: Respond to the search request initiated by the business terminal. The search request carries search keywords.

[0064] In different business scenarios, the business terminal will initiate different search requests, and the keywords used in each search request will also be different. Search keywords can be related terms determined based on the search input. For example, if the search word input from the business terminal is "water," the related term "fruit" will be determined from the related term library, and "fruit" will be used as the search keyword for the search request. Search keywords can also be input from the business terminal itself; the method of determining search keywords is not limited here.

[0065] Step 204: Obtain the business orders associated with the search keywords.

[0066] Specifically, the system retrieves the number of searches for related keywords. If the number of searches is less than a preset number, the process ends. If the number of searches is greater than the preset number, the system retrieves the business orders associated with the search keywords. The preset number of searches can be customized.

[0067] Step 206: Determine at least one first related term associated with the search keyword based on the business data value in the business order.

[0068] In this context, business data values ​​can be understood as the resources required for an item, such as the item's transaction value. The magnitude of the business data value characterizes the degree of relevance between related terms and search keywords. Each related term has a corresponding business data value; the larger the business data value, the higher the degree of relevance between the related term and the search keyword.

[0069] Specifically, the system uses a keyword search engine to retrieve business orders associated with the search keywords, identifies the items corresponding to the search keywords, aggregates and statistically analyzes the business orders associated with the search keywords based on different dimensions of the item's characteristics, and obtains business orders belonging to different item characteristics. Then, based on the business data values ​​in each business order, the business orders are sorted according to the item characteristics in a set order, resulting in a sorted order. At least one related term associated with the search keywords is obtained from the sorted business orders. Item characteristics include the item's category, brand, and attributes. The keyword search engine can be, but is not limited to, Solr. The keyword search engine can retrieve corresponding related items and their corresponding business data values ​​based on the keywords.

[0070] For example, if the search keyword is "apple," and there are items A, B, and C, the corresponding business orders are: Item A, category: mobile phone, business data value: 3000; Item B, category: fruit, business data value: 5000; Item C, category: mobile phone, business data value: 7000. Accumulating the business data values ​​for items in the mobile phone category, we get a business data value of 10000 for the mobile phone category and 5000 for the fruit category. As another example, if the search keyword is "fruit," the historical business orders for the fruit category include: apples (total transaction value: 5000), oranges (total transaction value: 4000), and bananas (total transaction value: 3000). Therefore, "fruit" yields three related category terms: apple, orange, and banana.

[0071] Step 208: Obtain the item click-through rate table that matches the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table.

[0072] The item click-through rate (CTR) table is determined based on historical click data. This involves analyzing click logs generated from historical searches, extracting the relationship between the search terms entered in the logs and the clicked items, and determining the CTR lookup table for at least one item feature. The CTR lookup table includes item feature fields and probability fields. Related terms in the item feature fields correspond to specific item CTRs; the higher the item CTR, the stronger the correlation between the related terms and the search keywords.

[0073] For example, the click log includes the correspondence between the search term "apple" and the clicked item categories. If the search term is "apple", there are click records for categories A, B, and C. From the click log, the number of clicks for different categories can be determined, and the click counts for categories A, B, and C can be obtained. Based on the click counts for categories A, B, and C, the click probabilities for categories A, B, and C can be determined respectively. The click probability can be understood as being determined based on the number of clicks for each category within a set time period and the total number of clicks for all categories.

[0074] Specifically, the search keyword is obtained, and a query is performed from a preset item click-through rate (CTR) table to obtain the item CTR table that matches the search keyword. Based on the item CTR table, at least one second related word associated with the search keyword is determined from the item CTR table.

[0075] Step 210: Based on the business data value of the first related word and the click rate of the second related word, determine the target related word of the search keyword from at least one first related word and at least one second related word.

[0076] The first related keyword determined based on business data values ​​and the second related keyword determined based on click-through rate can be the same or not exactly the same.

[0077] Specifically, when determining the target related words of the search keyword based on the first related word and the second related word, it is necessary to normalize the business data value of the first related word and the click rate of the second related word respectively to obtain the actual relatedness of each first related word, each second related word and the search keyword, and determine the target related words of the search keyword based on the actual relatedness of each first related word, each second related word and the search keyword.

[0078] In the aforementioned method for determining related terms, when determining related terms carrying search keywords, at least one first related term is determined based on business orders associated with the search keyword, and at least one second related term is determined from the item click-through rate (CTR) table. When determining related terms based on CTR, the corresponding item CTR table is further matched based on the search keyword. This allows for the determination of corresponding item CTR tables for different search keywords, further accurately determining related terms from the item CTR dimension. This avoids the situation where matching is limited to specific search keywords, which may not meet the needs of actual scenarios and lead to inaccurate second related terms. The target related terms for the search keyword are determined based on the business data value of the first related term and the CTR of the second related term. When determining the target related terms for the search keyword, and when determining related terms based on historical search keywords, the related terms for the search keyword are confirmed from both the business data value dimension and the item CTR dimension, avoiding the determination of target related terms based on a single dimension and improving the accuracy of related terms.

[0079] When determining the item click-through rate (CTR) table for search keywords, if the input search keywords are custom-input, directly matching the search keywords with the item CTR table will not find any items with CTRs related to the search keywords. However, there are relevant item click records in the actual historical click logs. Therefore, in order to improve the accuracy of related keywords, the search keywords can be split into words and matched.

[0080] In one embodiment, such as Figure 3 As shown, a method for determining the click-through rate (CTR) of items based on search keywords is provided, including the following steps:

[0081] Step 302: Obtain search keywords.

[0082] Step 304: Determine if there is a click-through rate table for items that match the search keywords. If yes, proceed to step 306; otherwise, proceed to step 308.

[0083] Understandably, the item click-through rate (CTR) table is determined based on historical click logs. This involves parsing the click logs to obtain the relationship between the input query terms and the item characteristics of the clicked items. Based on the query terms and item characteristics, a first item CTR table is determined. Taking item categories as an example, the correspondence between query terms and the categories of clicked items is obtained, the number of clicks for the corresponding category is determined, and the CTR for each category is obtained. Furthermore, when performing a search, the input keywords can be customized. To accurately determine the association, the query terms in the click logs can be split into segments, and the relationship between each segment and the item characteristics of the clicked items is determined. Based on the segmented query terms and item characteristics, a second item CTR table is determined. The splitting of query terms can be done according to actual needs and is not limited here.

[0084] Step 306: Obtain the click-through rate table of items matching the search keywords.

[0085] Step 308: If there is no item click-through rate table matching the search keyword, then the search keyword is split to obtain at least two sub-keywords.

[0086] Step 310: Determine the matching item click-through rate table from the preset database based on at least two sub-keywords.

[0087] In the above embodiments, when obtaining the item click-through rate table based on the search keywords, if there is no item click-through rate table matching the search keywords, the search keywords are split to obtain the matching item click-through rate table. This avoids determining the item click-through rate table matching the search keywords in a single way, and improves the accuracy and reliability of related terms.

[0088] Furthermore, in one embodiment, when a search request carrying search keywords is received, if the format of the search keywords does not conform to a preset format, the search keywords are converted to obtain search keywords in a preset format. Then, based on the search keywords in the preset format, the associated business orders can be accurately determined, the first associated term can be identified, and / or, the matching item click-through rate table can be accurately determined, obtaining the second associated term. In actual searches, due to different input methods on business terminals, the obtained search keywords may be characters or pinyin, etc. When the search keyword is pinyin, to accurately determine the associated business orders or item click-through rate tables, the pinyin can be converted to characters. Based on the converted characters, the associated business orders or item click-through rate tables can be accurately determined.

[0089] In one embodiment, such as Figure 4 As shown, a method for determining related keywords based on click-through rate is provided, and this method is applied to... Figure 1Taking the server in [ ] as an example, the following steps are included:

[0090] Step 402: Respond to the search request initiated by the service terminal, where the search request carries a search keyword.

[0091] Step 404: If there is no item click-through rate table that matches the search keyword, split the search keyword to obtain at least two sub-keywords.

[0092] Specifically, if there is no item click-through rate table that matches the search keyword in the preset database, split the search keyword to obtain at least two sub-keywords. Among them, splitting the search keyword can be done by splitting the search keyword according to a preset number of characters, or by performing semantic recognition on the search keyword and splitting it according to the recognized semantic result. The semantic recognition can be achieved through semantic recognition methods, which will not be elaborated here.

[0093] Step 406: Determine the matching item click-through rate table from the preset database according to at least two sub-keywords.

[0094] Step 408: Respectively determine the third related words associated with each sub-keyword from the item click-through rate table.

[0095] Specifically, according to the two sub-keywords obtained after splitting, determine the matching item click-through rate table from the preset database, and respectively determine the third related words associated with each sub-keyword in the item click-through rate table. <​​​​​​​​​​​​​​

[0101] Then: the probability score of category A = probability A1 + probability A3; the probability score of category B = probability B1 + probability B2 + probability B3; the probability score of category C = probability C2; the probability score of category D = probability D2; the probability scores of the four categories are normalized and mapped to the range [0,1].

[0102] In the above embodiments, when there is no item click-through rate table matching the search keyword in the preset database, at least two sub-keywords are obtained by splitting the search keyword. The matching item click-through rate table is determined based on each sub-keyword. The related words of each sub-keyword are weighted to obtain the correlation between the related words of each sub-keyword and the search keyword. The related words of the search keyword are further determined. By splitting the keyword, more matching data is used, thereby achieving more accurate matching and making the reliability of the related words higher.

[0103] In one embodiment, such as Figure 5 As shown, a method for determining business orders associated with search keywords is provided, which can be applied to... Figure 1 Taking the server in the example, the following steps are included:

[0104] Step 502: If there is no business order in the business order database associated with the search keyword, then extract the characters of the search field from each business order in the business order database.

[0105] Step 504: If the similarity between the characters in the search field and the search keyword reaches a preset value, and the pinyin of the characters in the search field is the same as that of the search keyword, then the business order corresponding to the search field is the business order associated with the search keyword.

[0106] In the above embodiments, when determining the business orders associated with the search keyword, if there are no business orders associated with the search keyword in the business order database, the characters of the search field in each business order in the business order database are extracted. Based on the similarity between the characters of the search field and the search keyword, as well as the pinyin of the characters of the search field, the number of matching data is increased from the business orders associated with the search keyword, which can achieve more accurate matching and make the reliability of the associated words higher.

[0107] In another embodiment, such as Figure 6 As shown, a method for determining conjunctions is provided, which can be applied to... Figure 1 Taking the server in the example, the following steps are included:

[0108] Step 602: Respond to the search request initiated by the business terminal. The search request carries search keywords.

[0109] Step 604: Obtain the business orders associated with the search keywords.

[0110] Step 606: Determine at least one first related term associated with the search keyword based on the business data value in the business order.

[0111] Step 608: Determine whether a click-through rate table for items matching the search keywords exists in the preset database. If it exists, proceed to step 610; otherwise, proceed to step 612.

[0112] Step 610: Obtain the item click-through rate table that matches the search keyword, and determine at least one second related word associated with the search keyword from the item click-through rate table.

[0113] Step 612: If there is no item click rate table matching the search keyword in the preset database, then the search keyword is split to obtain at least two sub-keywords.

[0114] Step 614: Determine the matching item click-through rate table from the preset database based on at least two sub-keywords.

[0115] Specifically, a click-through rate table for items that match at least two sub-keywords is determined from a pre-set database.

[0116] Step 616: Determine the third associated keyword for each sub-keyword from the second item click rate table.

[0117] Step 618: Perform weighted processing based on the click-through rate of each third related term, and determine at least one second related term that is associated with the search keyword from the third related terms.

[0118] Step 620: Calculate the association value of each associated word by weighting the business data value of the first associated word and the click-through rate of the second associated word.

[0119] Step 622: Sort at least one first related word and at least one second related word according to the association value, and determine the target related words of the search keyword in descending order of association value.

[0120] In the above embodiments, when determining related terms for search keywords, a first related term is determined based on business data values, and at least one second related term is determined based on item click-through rate (CTR). When determining related terms for search keywords based on CTR, the corresponding item CTR table is further matched based on the search keyword. This allows for the determination of corresponding item CTR tables for different search keywords, further accurately determining related terms from the item CTR dimension, thus increasing the number of matched related terms. When determining target related terms for search keywords, and when determining related terms for search keywords based on historical search keywords, the related terms for search keywords are confirmed from both the dimensions of business data values ​​and item CTR, avoiding the determination of target related terms based on a single dimension and improving the accuracy of related terms.

[0121] Optionally, in one embodiment, when determining the first keyword associated with the search keyword, the search keyword composed of the search keyword and special characters is determined by obtaining matching target historical search fields from historical search fields. Each target historical search field corresponds to a business order containing business data values. Based on the business data values, the first candidate related words associated with the search keyword are extracted from the target historical search fields. The search keyword composed of the search keyword and special characters can be determined based on search behavior. For example, a search keyword table is queried for search keywords starting with the keyword plus a space. For each retrieved keyword, the word after the space is extracted as a related word. For example, if the search keyword is "fruit", the historical search fields are queried using "fruit" + a space, i.e., "fruit". For example, the search keyword table is searched, and "fresh fruit" is found. The word "fresh" is extracted and used as a related word. The business data values ​​of the search results for "fresh fruit" are then queried and statistically analyzed. Suppose the business data value of the search result for "fresh fruit" is 3000. "Fruit" has the related word "fresh". For example, if "fruit" + space queries the historical search field and get "fresh fruit in region A", then "fresh" and "region A" are used as related words for "fruit", and the business data values ​​of "fresh fruit" and "fruit in region A" are calculated separately.

[0122] The system identifies business orders associated with the search keyword from the business order database. Based on the business data values ​​within these orders, it determines second candidate related terms. These first and second candidate related terms are then ranked according to their respective business data values ​​to determine at least one first related term associated with the search keyword. By using search keywords and special characters, as well as directly determining related terms based on the search keyword format, the amount of matching data is increased, thus improving the accuracy of the related terms.

[0123] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0124] Based on the same inventive concept, this application also provides a device for determining related words to implement the method for determining related words described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the device for determining related words provided below can be found in the limitations of the method for determining related words above, and will not be repeated here.

[0125] In one embodiment, such as Figure 7 As shown, a device for determining related terms is provided, comprising: a response module 702, an order acquisition module 704, a first determining module 706, a second determining module 708, and a third determining module 710, wherein:

[0126] The response module 702 is used to respond to search requests initiated by the business terminal, and the search requests carry search keywords;

[0127] The order retrieval module 704 is used to retrieve business orders associated with search keywords.

[0128] The first determining module 706 is used to determine at least one first associated word associated with the search keyword based on the business data value in the business order.

[0129] The second determining module 708 is used to obtain an item click-through rate table that matches the search keyword, and to determine at least one second related word associated with the search keyword from the item click-through rate table.

[0130] The third determining module 710 is used to determine the target related words of the search keyword from at least one first related word and at least one second related word based on the business data value of the first related word and the click rate of the second related word.

[0131] In the aforementioned related term determination device, when determining related terms carrying search keywords, at least one first related term associated with the search keyword is determined based on business orders associated with the search keyword, and at least one second related term associated with the search keyword is determined from an item click-through rate (CTR) table. When determining related terms based on CTR, the corresponding item CTR table is further matched based on the search keyword. This allows for the determination of corresponding item CTR tables for different search keywords, further accurately determining related terms associated with the search keyword from the item CTR dimension. This avoids the situation where matching is limited to specific search keywords, which may not meet the needs of actual scenarios and lead to inaccurate determination of the second related term. The target related term for the search keyword is determined based on the business data value of the first related term and the CTR of the second related term. When determining the target related term for the search keyword, and when determining related terms based on historical search keywords, the related terms for the search keyword are confirmed from both the business data value dimension and the item CTR dimension, avoiding the determination of the target related term based on a single dimension and improving the accuracy of the related terms.

[0132] In another embodiment, a device for determining related terms is provided, which, in addition to including a response module 702, an order acquisition module 704, a first determination module 706, a second determination module 708, and a third determination module 710, further includes: a word splitting module, a weighted processing module, a character extraction module, an order acquisition module, and a format conversion module, wherein:

[0133] The keyword splitting module is used to split the search keyword into at least two sub-keywords if there is no item click-through rate table matching the search keyword in the preset database.

[0134] The second determining module 708 is also used to determine the matching item click rate table from a preset database based on at least two sub-keywords.

[0135] The second determining module 708 is also used to determine the third associated words associated with each sub-keyword from the item click rate table.

[0136] The weighted processing module is used to perform weighted processing based on the click-through rate of each third related term, and to determine at least one second related term that is associated with the search keyword from the third related terms.

[0137] The character extraction module is used to extract characters from the search field of each business order in the business order database if there is no business order in the business order database associated with the search keyword.

[0138] The order acquisition module is also used to assign the business order corresponding to the search field to the business order associated with the search keyword if the similarity between the characters in the search field and the search keyword reaches a preset value, and the pinyin of the characters in the search field is the same as the search keyword.

[0139] The weighted processing module is used to perform weighted calculations based on the business data value of the first associated word and the click rate of the second associated word, so as to obtain the association value of each associated word.

[0140] The third determining module 710 is also used to sort at least one first related word and at least one second related word according to the association value, and determine the target related words of the search keyword in descending order of association value.

[0141] The format conversion module is used to obtain the format of the search keywords; if the format of the search keywords does not conform to the preset format, the format of the search keywords is converted to obtain the search keywords in the preset format.

[0142] Each module in the aforementioned device for determining related terms can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0143] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores search-related data such as search data, business orders, click logs, and item click-through rate tables. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a method for determining related terms.

[0144] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0145] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0146] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0147] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0148] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0149] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0150] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method of determining a conjunction word, characterized by, The method comprises: In response to a search request initiated by a service terminal, the search request carrying a search keyword; Obtain the service order associated with the search keyword; the service order is obtained by aggregating and counting the service orders associated with the search keyword according to the multi-dimensional characteristics of the goods corresponding to the search keyword, and the service orders belonging to different goods characteristics are obtained; Determine at least one first associated word associated with the search keyword according to the business data value in the service order; the first associated word is determined based on the ranking of the first candidate associated word and the second candidate associated word, the first candidate associated word is obtained by combining the search keyword and the special character to query the historical search field, and the target historical search field is matched, and the associated word related to the search keyword is extracted from the target historical search field according to the business data value corresponding to the target historical search field, and the second candidate associated word is determined according to the business data value in the service order and the associated word associated with the search keyword; Obtain the item click rate table matched with the search keyword, and determine at least one second associated word associated with the search keyword from the item click rate table; Determine the target associated word of the search keyword from the at least one first associated word and the at least one second associated word according to the business data value of the first associated word and the click rate of the second associated word.

2. The method of claim 1, wherein, The method comprises: If there is no item click rate table matched with the search keyword in the preset database, split the search keyword to obtain at least two sub keywords; Determine the matched item click rate table from the preset database according to the at least two sub keywords.

3. The method of claim 2, wherein, The method comprises: Determine the third associated word associated with each of the sub keywords from the item click rate table respectively; According to the click rate of each of the third associated words, at least one second associated word associated with the search keyword is determined from the third associated words.

4. The method of claim 1, wherein, The method comprises: If there is no service order associated with the search keyword in the service order library, extract the characters of the search field in each service order in the service order library; If the similarity of the characters of the search field and the search keyword reaches a preset value, and the pinyin of the characters of the search field is the same as the search keyword, the service order corresponding to the search field is the service order associated with the search keyword.

5. The method of claim 1, wherein, The method comprises: After normalizing and weighting the business data value of the first associated word and the click rate of the second associated word, the associated value of each associated word is obtained respectively. According to the association values, the at least one first association word and the at least one second association word are sorted in descending order of the association values, and a target association word of the search keyword is determined.

6. The method of claim 1, wherein, Before the business order associated with the search keyword is obtained, the method further comprises: Obtaining the format of the search keyword; If the format of the search keyword does not conform to the preset format, the format of the search keyword is converted to obtain a search keyword in the preset format.

7. An apparatus for determining a conjunction word, characterized by comprising: The device comprises: A response module for responding to a search request initiated by a business terminal, the search request carrying a search keyword; An order obtaining module for obtaining a business order associated with the search keyword; the business order is obtained by aggregating and counting business orders associated with the search keyword according to multi-dimensional features of an item corresponding to the search keyword, and belongs to different item features; A first determining module for determining at least one first association word associated with the search keyword according to a business data value in the business order; the first association word is determined based on the sorting of first candidate association words and second candidate association words, the first candidate association word is obtained by combining the search keyword and a special character to query a historical search field, extracting a target historical search field matching the search keyword, and extracting a first association word related to the search keyword from the target historical search field according to a business data value corresponding to the target historical search field, and the second candidate association word is an association word determined according to the business data value in the business order; A second determining module for obtaining an item click rate table matching the search keyword, and determining at least one second association word associated with the search keyword from the item click rate table; A third determining module for determining a target association word of the search keyword from the at least one first association word and the at least one second association word according to the business data value of the first association word and the click rate of the second association word.

8. The apparatus of claim 7, wherein, The device further comprises: A word splitting module for splitting the search keyword to obtain at least two sub-keywords if the item click rate table matching the search keyword does not exist in the preset database; The second determining module is further configured to determine a matching item click rate table from the preset database according to the at least two sub-keywords.

9. The apparatus of claim 8, wherein, The second determining module is further configured to determine a third association word associated with each of the sub-keywords from the item click rate table. The device further comprises: A weighting processing module for weighting processing according to the click rates of the third association words, and determining at least one second association word associated with the search keyword from the third association words.

10. The apparatus of claim 7, wherein, The device further comprises: A character extracting module for extracting characters of a search field in each business order in the business order library if the business order associated with the search keyword does not exist in the business order library. The order obtaining module is further configured to, if the similarity between the character of the search field and the search keyword reaches a preset value and the pinyin of the character of the search field is the same as the search keyword, determine the business order corresponding to the search field as the business order associated with the search keyword.

11. The apparatus of claim 7, wherein, The apparatus further includes: The weighting processing module is configured to normalize and weight the business data value of the first associated word and the click rate of the second associated word to obtain the associated value of each associated word. The third determining module is further configured to sort the at least one first associated word and the at least one second associated word according to the associated values, and determine the target associated word of the search keyword in descending order of the associated values.

12. The apparatus of claim 7, wherein, The apparatus further includes: The format conversion module is configured to obtain the format of the search keyword. If the format of the search keyword does not conform to the preset format, the format of the search keyword is converted to obtain the search keyword in the preset format.

13. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to implement the steps of the method in any one of claims 1 to 6.

14. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6.

15. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6.

Images