Shop matching method and device and computer equipment

Abstract

The invention provides a shop matching method, a shop matching device and computer equipment, and the method comprises the steps: receiving question information, analyzing the question information, and obtaining a corresponding initial keyword; determining candidate keywords, wherein the semantic similarity between the candidate keywords and the initial keyword is greater than or equal to a first threshold value; combining the initial keyword with the candidate keyword to obtain a target keyword; and determining a preset shop matched with the question information according to a preset shop label and the target keyword. According to the invention, the accuracy and efficiency of shop matching can be improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, relates to intelligent question-answering technology, and in particular to a store matching method, device and computer equipment. Background Art

[0002] In the diverse scenarios of providing product services to users (for example, services at different stages such as pre-sales service, sales service, and after-sales service), users often ask various questions, and it is necessary to help users solve their problems more quickly and improve user experience. However, when actually answering user questions, it is often limited by the professional knowledge and experience of the question answerers, which can easily affect the accuracy of the answer information and lead to low answer efficiency.

[0003] For example, in an e-commerce scenario, when a user inquires about a store's relevant information (e.g., a store's success stories, sales product information, etc.), if the business personnel only rely on personal experience, it is difficult to provide effective information accurately and quickly, resulting in low accuracy and efficiency of store matching. Summary of the invention

[0004] The embodiments of the present application provide a store matching method, apparatus and computer equipment to solve the problems of low accuracy and low efficiency of store matching.

[0005] In a first aspect, an embodiment of the present application provides a store matching method, which includes: receiving question information, parsing the question information, and obtaining corresponding initial keywords; determining candidate keywords whose semantic similarity with the initial keywords is greater than or equal to a first threshold; combining the initial keywords with the candidate keywords to obtain target keywords; and determining a preset store that matches the question information based on a preset store tag and the target keyword.

[0006] In a feasible embodiment, before receiving question information, parsing the question information, and obtaining corresponding initial keywords, the method also includes: determining a problem description corresponding to the question information; determining a problem classification corresponding to the question information based on a correspondence between the problem description and the problem classification; if the problem classification is a target category, parsing the question information to obtain the initial keywords.

[0007] In a feasible embodiment, parsing the question information to obtain the corresponding initial keyword includes: performing word segmentation processing on the question information to obtain multiple word segments; determining the word segmentation attribute corresponding to each word segment; and using the word segmentation corresponding to the target word segmentation attribute as the initial keyword.

[0008] In a feasible embodiment, determining the candidate keywords whose semantic similarity with the initial keywords is greater than or equal to a first threshold includes: based on a preset correspondence between the initial keywords and the candidate keywords, determining the candidate keywords whose semantic similarity with the initial keywords is greater than or equal to the first threshold.

[0009] In a feasible embodiment, the method also includes: determining the business category and business products corresponding to the preset store; determining the first label of the preset store based on the business category, and determining the second label of the preset store based on the business products; combining the first label and the second label to obtain the store label.

[0010] In a feasible embodiment, determining the preset store that matches the question information based on the preset store label and the target keyword includes: determining the preset store corresponding to the store label that is the same as any target keyword; determining the number of matching preset stores; if the number of stores is less than a second threshold, performing semantic matching based on the store label and the target keyword to obtain a semantic matching result; determining the preset store that matches the question information based on the semantic matching result.

[0011] In a feasible embodiment, semantic matching is performed based on the store tag and the target keyword to obtain a semantic matching result, including: determining a first vector corresponding to the store tag; determining a second vector corresponding to the target keyword; if the similarity between the first vector and the second vector is greater than a third threshold, determining that the semantic matching result is a semantic match between the store tag and the target keyword; if the similarity between the first vector and the second vector is less than or equal to the third threshold, determining that the semantic matching result is a semantic mismatch between the store tag and the target keyword.

[0012] In a feasible embodiment, the method further includes: if the number of stores is greater than or equal to the second threshold, sorting the matching preset stores according to preset rules; and selecting a preset number of preset stores ranked first as preset stores matching the question information.

[0013] In a second aspect, an embodiment of the present application provides a store matching device, which includes: a question receiving module, used to receive question information, parse the question information, and obtain corresponding initial keywords; a candidate determination module, used to determine candidate keywords whose semantic similarity with the initial keywords is greater than or equal to a first threshold; a keyword combination module, used to combine the initial keywords with the candidate keywords to obtain target keywords; a store matching module, used to determine a preset store that matches the question information based on a preset store tag and the target keyword.

[0014] In a third aspect, an embodiment of the present application provides a computer device, comprising a controller and a memory, wherein the controller is configured to implement any one of the above-described store matching methods when executing a computer program stored in the memory.

[0015] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a controller, any one of the above-mentioned store matching methods is implemented.

[0016] The embodiment of the present application provides a store matching method, which receives question information, parses the question information, obtains the corresponding initial keyword; determines the candidate keyword whose semantic similarity with the initial keyword is greater than or equal to a first threshold; combines the initial keyword with the candidate keyword to obtain the target keyword; and determines the preset store that matches the question information based on the preset store label and the target keyword. The above method can improve the accuracy of store matching by expanding the initial keyword in the question information to obtain the candidate keyword with semantic similarity to the initial keyword, and using the initial keyword and the candidate keyword to determine the preset store corresponding to the question information; and the present application determines the preset store corresponding to the question information by matching the target keyword with the store label of the preset store, thereby realizing automatic determination of the preset store corresponding to the question information and improving the efficiency of store matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of a computer device provided in an embodiment of the present application.

[0018] Figure 2 This is a flow chart of a store matching method provided in an embodiment of the present application.

[0019] Figure 3 It is a flowchart of the intent recognition method provided in an embodiment of the present application.

[0020] Figure 4 It is a flowchart of the initial keyword determination method provided in an embodiment of the present application.

[0021] Figure 5 It is a flow chart of the store label determination method provided in an embodiment of the present application.

[0022] Figure 6 It is a flowchart of the preset store determination method provided in an embodiment of the present application.

[0023] Figure 7 It is a structural diagram of a store matching device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] It should be noted that the terms "first" and "second" in the specification, claims and drawings of the present application are used to distinguish similar objects rather than to describe a specific order or sequence.

[0025] It should also be noted that the method disclosed in the embodiments of the present application or the method shown in the flowchart includes one or more steps for implementing the method. Without departing from the scope of the claims, the execution order of multiple steps can be interchangeable with each other, and some of the steps can also be deleted.

[0026] Some embodiments will be described below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments may be combined with each other.

[0027] See also Figure 1 , Figure 1 is a schematic diagram of a computer device provided in an embodiment of the present application. As an implementation mode, the store matching method provided in an embodiment of the present application can be applied to a computer device, for example, Figure 1 The computer device 10 shown. The computer device 10 can be a server cluster or a single server. Through the computer device 10, multiple independent stations can be built. The number of independent stations can be set according to actual needs, and multiple preset stores can be established in each independent station.

[0028] In some embodiments, the computer device 10 includes a memory 11, at least one controller 12, and at least one communication bus 13. The controller 12 is used to execute the computer program stored in the memory 11 to implement the store matching method provided in the embodiment of the present application. The at least one communication bus 13 is configured to realize the connection and communication between the memory 11 and the at least one controller 12.

[0029] Figure 1 The structure of the computer device shown does not constitute a limitation on the embodiments of the present application. The computer device 10 may also include more or less other hardware or software than shown in the figure, or a different arrangement of components.

[0030] In some embodiments of the present application, the computer device 10 can also be connected to a client device, which includes but is not limited to any electronic product that can interact with a user through a keyboard, mouse, remote control, touchpad or voice control device, such as a personal computer, a tablet computer, a smart phone, a digital camera, etc.

[0031] It should be noted that the computer device 10 is only an example, and other existing or future electronic products that are suitable for the present application should also be included in the protection scope of the present application and included here by reference.

[0032] In some embodiments, the computer device 10 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0033] The present application will describe the technical solution of the present application in detail through specific embodiments below. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0034] Figure 2 is a flow chart of a store matching method provided in an embodiment of the present application, the store matching method is applied to a computer device (for example, Figure 1 Computer device 10 in FIG. Figure 2 As shown, the store matching method may include the following steps S11-S14. According to different requirements, the order of the steps in the flowchart may be changed, and some may be omitted.

[0035] S11, receiving question information, parsing the question information, and obtaining corresponding initial keywords.

[0036] In at least one embodiment of the present application, the question information may represent consulting information related to the store input by the user. For example, the question information may include "Are there any successful store case references related to solar energy?", "Are there any successful store case references related to bird's nest?", etc.

[0037] In some embodiments, the question information may include one or more segmentation words, and each segmentation word has a corresponding segmentation attribute. The initial keyword may represent a segmentation word whose segmentation attribute in the question information is a target segmentation attribute. Among them, the target segmentation attribute may be a product attribute. Continuing with the above embodiment, when the question information is "Are there any successful store case references related to solar energy", the segmentation attribute corresponding to "solar energy" is a product attribute, so the initial keyword is "solar energy"; when the question information is "Are there any successful store case references related to bird's nest", the segmentation attribute corresponding to "bird's nest" is a product attribute, so the initial keyword is "bird's nest".

[0038] S12, determining candidate keywords whose semantic similarity with the initial keyword is greater than or equal to a first threshold.

[0039] In at least one embodiment of the present application, the initial keyword is semantically expanded to obtain multiple candidate keywords whose semantic similarity with the initial keyword is greater than or equal to a first threshold, wherein the number of candidate keywords can be set according to actual needs, for example, the number of candidate keywords can be one, two, three, etc. The first threshold can be set according to actual needs and is not limited here.

[0040] In some embodiments, a keyword library can be pre-set, and the keyword library contains multiple semantically similar words. According to the initial keyword, the keyword library is queried to obtain candidate keywords with semantics similar to the initial keyword. In other embodiments, a synonym determination model can also be pre-trained, and the input data of the synonym determination model is the initial keyword, and the output data is the candidate keyword with semantics similar to the initial keyword. The synonym determination model may include but is not limited to convolutional neural network models such as LeNet-5, AlexNet, VGGNet, ResNet, Inception, or supervised learning models such as support vector machines. The above examples of models are only examples and are not limited to this in actual applications.

[0041] Continuing with the above embodiment, when the initial keyword is "solar energy", the candidate keywords may include solar thermal energy, light energy, solar radiation energy, solar energy, solar energy, etc.; when the initial keyword is "bird's nest", the candidate keywords may include bird's nest, bird's nest root and bird's nest vegetables, without any limitation.

[0042] S13, combining the initial keyword and the candidate keyword to obtain a target keyword.

[0043] In at least one embodiment of the present application, the initial keyword is combined with the candidate keyword to obtain a target keyword, and the number of target keywords can be multiple. In some embodiments, if there is no candidate keyword whose semantic similarity with the initial keyword is greater than or equal to the first threshold, the initial keyword is used as the target keyword.

[0044] In some embodiments, the combination of the initial keyword and the candidate keyword to obtain the target keyword includes: determining a first frequency corresponding to the initial keyword and a second frequency corresponding to the candidate keyword; determining the sorting information of the initial keyword and the candidate keyword based on the first frequency and the second frequency; combining the initial keyword and the candidate keyword based on the sorting information to obtain the target keyword. Among them, the first frequency can represent the frequency of use of the initial keyword, and the second frequency can represent the frequency of use of the candidate keyword. According to the first frequency and the second frequency, the initial keyword and the candidate keyword are sorted based on a preset sorting of frequencies (for example, from high to low) to obtain the target keyword. The embodiment of the present application sorts the initial keyword and the candidate keyword based on the frequency of use, so that the keywords with higher frequency of use are sorted first, and the efficiency of store matching can be improved when matching is performed based on the target keyword and the preset store in the subsequent process.

[0045] S14, determining a preset store matching the question information according to the preset store tag and the target keyword.

[0046] In at least one embodiment of the present application, a store tag is used to indicate a category tag and a product tag of a preset store, wherein the category tag is used to indicate the business category of the preset store, and the product tag is used to indicate the product operated by the preset store, and the category level to which the product belongs is greater than the preset level. Among them, the preset classification system includes categories within (including) the preset level. The preset level can be set according to actual classification needs and is not limited here. For example, taking the main category as 3C category and the preset level as the third-level category, the 3C category is taken as the first-level category, mobile phones, computers, photography or video equipment, household appliances, home office equipment, and games and entertainment equipment are taken as second-level categories, various brands and models of smart phones and mobile phone accessories are taken as the third-level category of mobile phones, laptops and desktop computers are taken as the third-level category of computers, cameras used for professional or amateur photography are taken as the third-level category of photography or video equipment, televisions, audio systems, air conditioners, washing machines, refrigerators, etc. are taken as the third-level category of household appliances, projectors, copiers, printers, etc. are taken as the third-level category of home office equipment, and tablet computers and e-book readers are taken as the third-level category of games and entertainment equipment. Take the preset store A operating smart phones, charging cables and charging heads as an example for explanation. Based on the above classification system, it can be determined that the category label corresponding to the preset store A is 3C category-mobile phone-smart phones and mobile phone accessories. Since the preset store A also sells charging cables and charging heads, charging cables and charging heads belong to the next level category of mobile phone accessories (that is, the fourth level category), and the fourth level category is not included in the above classification system, based on this, charging cables and charging heads are used as product labels for the preset store A.

[0047] In some embodiments, when matching the preset store tag with the target keyword, a combination of direct matching rules and semantic matching rules can be used. First, the preset store tag is directly matched with the target keyword. If the direct matching cannot meet the preset matching requirements, the preset store tag is semantically matched with the target keyword, and the preset store matching the question information is determined based on the semantic matching result; if the direct matching meets the preset matching requirements, the preset store matching the question information is determined based on the direct matching results. The preset matching requirements can be set according to actual needs and are not limited here.

[0048] An embodiment of the present application provides a store matching method, which semantically expands the initial keywords in the question information to obtain candidate keywords with similar semantics to the initial keywords, and uses the initial keywords and the candidate keywords to determine the preset store corresponding to the question information, thereby improving the accuracy of store matching; and the present application determines the preset store corresponding to the question information by matching the target keyword with the store label of the preset store, thereby realizing automatic determination of the preset store corresponding to the question information and improving the efficiency of store matching.

[0049] In at least one embodiment of the present application, before parsing the question information to obtain the corresponding initial keywords, the question information needs to be identified for intent, and when it is determined that the question information is of a specified intent, the question information is parsed to obtain the corresponding initial keywords. Figure 3 is a flow chart of the intention recognition method provided in the embodiment of the present application, and the intention recognition method is applied to a computer device. Figure 3 As shown, the following steps are included:

[0050] S21, determining the question description corresponding to the question information.

[0051] In at least one embodiment of the present application, the problem description may be a pre-set description template. For example, the problem description may include "Are there any store case references about **", "Are there any ** store cases", "Provide successful cases of ** type stores", "Establishment time of ** online store", "Monthly sales volume of ** store", etc.

[0052] In some embodiments, by semantically matching the question information with the preset question description, a question description with a semantic similarity greater than a preset threshold is selected as the question description corresponding to the question information. The preset threshold can be set according to actual needs and is not limited here.

[0053] S22, determining the question category corresponding to the question information according to the corresponding relationship between the question description and the question category.

[0054] In at least one embodiment of the present application, a correspondence is established between the question description and the question classification. The question classification may include store questions and non-store questions. By querying the correspondence between the question description and the question classification, the question classification corresponding to the question information can be determined.

[0055] Continuing with the above embodiment, the problem description includes "Are there any references to store cases about **", "Are there any ** store cases", "Provide successful cases of ** type stores", "Establishment time of ** online store", "Monthly sales volume of ** store", among which, the problem categories corresponding to the three question descriptions "Are there any references to store cases about **", "Are there any ** store cases", and "Provide successful cases of ** type stores" are store problems, and the problem categories corresponding to the two question descriptions "Establishment time of ** online store" and "Monthly sales volume of ** store" are non-store problems.

[0056] S23: If the question is classified as a target category, the question information is parsed to obtain the initial keywords.

[0057] In at least one embodiment of the present application, the target category may be a store question. If the question corresponding to the question information is classified as a store question and the question information is determined to be of a specified intent, the question information is parsed to obtain the initial keywords. If the question corresponding to the question information is classified as a non-store question and the question information is determined not to be of a specified intent, a prompt message is output, and the prompt message can be used to prompt the user that the current question information is a non-store question.

[0058] In the store matching method provided in the embodiment of the present application, by determining the question description corresponding to the question information, and determining the question classification corresponding to the question information based on the question description, when the question classification is the target category, the question information is parsed to obtain the initial keywords. The above method can screen out question information with non-specified intent by identifying the intent of the question information, thereby improving the efficiency of store matching.

[0059] Figure 4 1 is a flow chart of an initial keyword determination method provided in an embodiment of the present application, and the initial keyword determination method is applied to a computer device. Figure 4 As shown, the following steps are included:

[0060] S31, performing word segmentation processing on the question information to obtain a plurality of word segments.

[0061] In at least one embodiment of the present application, word segmentation processing is the process of dividing the continuous text in the question information into independent words, and the question information contains one or more word segments. In some embodiments, a word segmentation system can be pre-set, and the word segmentation system contains multiple commonly used words and their corresponding parts of speech. Based on the word segmentation system, the question information can be word segmented to obtain multiple word segments and the part of speech of each word segment. In other embodiments, word segmentation processing can also be performed by deep learning, for example, using a recurrent neural network or a long short-term memory network to model sequence data, and word segmentation processing is performed by capturing the context information of the question information to obtain multiple word segments.

[0062] S32, determining the segmentation attribute corresponding to each segmentation.

[0063] In at least one embodiment of the present application, each participle has a corresponding part of speech, which may include verbs, adverbs, prepositions, adjectives, conjunctions, and nouns, etc., wherein verbs are words that express actions or states, such as eating, running, flying, thinking, swimming, etc. Adverbs are used to modify verbs, adjectives, or other adverbs to express time, place, manner, degree, etc. For example: quickly, here, very good, often, very, etc. Prepositions are used before nouns or pronouns to express their relationship with other words. For example: in..., about, to..., from..., for, etc. Adjectives are used to modify nouns to express properties, characteristics, or states. For example: happy, big, beautiful, smart, new, etc. Conjunctions are used to connect words, phrases, or sentences. For example: and, but, or, because, so, etc.; nouns represent the names of people, things, places, or concepts. For example: apple, book, city, happiness, freedom.

[0064] In some embodiments, the segmentation attributes may include product attributes, store attributes, etc. Based on the part of speech, a segmentation whose part of speech is a noun is selected, and it is determined whether the segmentation is a target segmentation attribute. In some embodiments of the present application, the target segmentation attribute may be a product attribute, and in other embodiments, the target segmentation attribute may also be a store attribute. Continuing with the above embodiment, when the question information is "Are there any successful store case references related to solar energy", the segmentations whose part of speech is a noun include "solar energy" and "store case", among which the segmentation attribute corresponding to "solar energy" is a product attribute, and the segmentation attribute corresponding to "store case" is a store attribute.

[0065] S33, taking the segmentation corresponding to the target segmentation attribute as the initial keyword.

[0066] In at least one embodiment of the present application, the initial keyword may represent a segmentation in the question information whose segmentation attribute is the target segmentation attribute. Continuing with the above embodiment, when the question information is "Are there any successful store case references related to solar energy?", the initial keyword is "solar energy".

[0067] In the store matching method provided in the embodiment of the present application, by performing word segmentation processing on the question information, determining the initial keywords related to the store in the question information, and determining the preset store corresponding to the question information based on the initial keywords, the accuracy of store matching can be improved.

[0068] In at least one embodiment of the present application, the initial keyword is semantically expanded to obtain multiple candidate keywords. In some embodiments, the determination of the candidate keywords whose semantic similarity with the initial keyword is greater than or equal to the first threshold comprises: based on a pre-set correspondence between the initial keyword and the candidate keywords, determining the candidate keywords whose semantic similarity with the initial keyword is greater than or equal to the first threshold. In some embodiments, a keyword library is pre-established, and the keyword library contains multiple keywords whose semantic similarity is greater than or equal to the first threshold. By traversing the keyword library based on the initial keyword, the candidate keywords whose semantic similarity with the initial keyword is greater than or equal to the first threshold can be determined, and the number of candidate keywords can be one or more, which is not limited here. The embodiment of the present application can improve the accuracy of store matching by semantically expanding the initial keyword and using the initial keyword and the candidate keyword to determine the preset store corresponding to the question information.

[0069] In at least one embodiment of the present application, a store tag of a preset store is predetermined to facilitate subsequent determination of a preset store matching the question information based on the store tag and the target keyword. Figure 5 is a flow chart of a method for determining a store label provided in an embodiment of the present application, and the method for determining a store label is applied to a computer device. Figure 5 As shown, the following steps are included:

[0070] S41, determining the business categories and business commodities corresponding to the preset store.

[0071] In at least one embodiment of the present application, the business category may describe the category of business activities that the preset store engages in. The business category may include category information of a preset level. The preset level may be set according to actual needs. For example, the preset level may include three levels of categories, and the corresponding business category may include three levels of business categories. The business commodity may describe the specific product operated by the preset store, and the category level corresponding to the product is greater than the preset level. Continuing from the above embodiment, the business category corresponding to the preset store may be the 3C category - mobile phones - smart phones and mobile phone accessories, and the business commodity may be charging cables and charging heads.

[0072] In some embodiments, the business category can be input by the merchant, or it can be determined by calculating the category main business rate for a preset store, wherein the category main business rate can be determined based on parameters such as the category transaction volume in the preset store in the past 30 days, the transaction volume of the preset store in the past 30 days, the number of online products in the category on the day, and the number of online products in the preset store on the day.

[0073] In some embodiments, the business commodity can be input by the merchant, and then the preset commodity classification model is used to determine whether the commodity input by the merchant is a business commodity. The commodity classification model can be a binary classification model. If the output result of the commodity classification model is a first preset value (e.g., 1), the commodity input by the merchant is determined to be a business commodity; if the output result of the commodity classification model is a second preset value (e.g., 0), the business category corresponding to the preset store can be determined based on the commodity input by the merchant.

[0074] S42, determining a first label of the preset store according to the business category, and determining a second label of the preset store according to the business products.

[0075] In at least one embodiment of the present application, a first label is constructed for the business category, and a second label is constructed for the business commodity, wherein the business category corresponding to the preset store can be determined by parsing the first label, and the business commodity corresponding to the preset store can be determined by parsing the second label. The method of establishing labels can refer to the relevant technology and will not be described in detail here.

[0076] S43: Combine the first tag and the second tag to obtain the store tag.

[0077] In the store matching method provided in the embodiment of the present application, by determining the business category and business products of the preset store, determining the first label of the preset store based on the business category, and determining the second label of the preset store based on the business products, the first label and the second label are used as store labels of the preset store, and matching is performed based on the store labels and the target keywords of the question information, thereby improving the accuracy of matching with the question information.

[0078] In at least one embodiment of the present application, when matching a preset store tag with a target keyword, a combination of direct matching rules and semantic matching rules can be used. First, the preset store tag is directly matched with the target keyword. If the direct matching cannot meet the preset matching requirements, the preset store tag is semantically matched with the target keyword, and the preset store that matches the question information is determined based on the semantic matching result; if the direct matching meets the preset matching requirements, the preset store that matches the question information is determined based on the direct matching result. Figure 6 1 is a flow chart of a method for determining a preset store provided in an embodiment of the present application, and the method for determining a preset store is applied to a computer device. Figure 6 As shown, the following steps are included:

[0079] S51, determining a preset store corresponding to a store tag identical to any target keyword.

[0080] In at least one embodiment of the present application, the store label of each preset store is matched with any target keyword among multiple target keywords, and the preset store corresponding to the store label that is the same as any target keyword is selected. The number of preset stores can be one or more.

[0081] S52: Determine the number of matching preset stores.

[0082] In at least one embodiment of the present application, the number of preset stores corresponding to the store tags identical to any target keyword is used as the number of stores.

[0083] S53: If the number of stores is less than a second threshold, semantic matching is performed based on the store tags and the target keywords to obtain a semantic matching result.

[0084] In at least one embodiment of the present application, if the number of stores is less than the second threshold, it indicates that direct matching cannot meet the preset matching requirements, and it is necessary to semantically match the preset store tags with the target keywords to obtain a semantic matching result. The second threshold can be set according to actual needs, for example, the second threshold can be three.

[0085] In some embodiments, if the number of stores is greater than or equal to the second threshold, the matching preset stores are sorted according to a preset rule; and the preset stores with the preset number of stores ranked first are selected as the preset stores corresponding to the question information. The preset rule can be set according to actual needs, for example, the preset rule can be sorted in order from high to low in terms of sales volume of the goods. The preset number can be set according to actual needs, for example, the preset number can be three.

[0086] S54: Determine a preset store that matches the question information according to the semantic matching result.

[0087] In at least one embodiment of the present application, the semantic matching result may include that the store tag does not semantically match the target keyword and that the store tag semantically matches the target keyword. When the semantic matching result is that the store tag does not semantically match the target keyword, a prompt message may be output to prompt the business personnel that no preset store matching the user's question information has been found. When the store tag semantically matches the target keyword, the number of preset stores matching the target keyword may be determined. If the number of preset stores matching the target keyword is greater than or equal to the second threshold, the matching preset stores are sorted according to the preset rules; and the preset number of preset stores ranked first is selected as the preset stores corresponding to the question information. Among them, the preset rules may be set according to actual needs, for example, the preset rules may be sorted in order from high to low according to the sales volume of the goods. The preset number may be set according to actual needs, for example, the preset number may be three. If the number of preset stores matching the target keyword is less than the second threshold, a prompt message is output to prompt the business personnel that no preset store matching the user's question information has been found.

[0088] In the store matching method provided in the embodiment of the present application, a combination of direct matching rules and semantic matching rules is used to determine a preset store that matches the question information, which can improve the accuracy of store matching.

[0089] In some embodiments, the semantic matching is performed based on the store tag and the target keyword to obtain a semantic matching result, including: determining a first vector corresponding to the store tag; determining a second vector corresponding to the target keyword; if the similarity between the first vector and the second vector is greater than a third threshold, then determining that the semantic matching result is a semantic match between the store tag and the target keyword; if the similarity between the first vector and the second vector is less than or equal to the third threshold, then determining that the semantic matching result is a semantic mismatch between the store tag and the target keyword. Among them, the text information can be represented as a vector that can express the semantics of the text based on word embedding technologies such as Word2Vec and Doc2Vec. The similarity of vectors can be measured in a variety of ways, including Euclidean distance, Manhattan distance, cosine similarity, Jaccard similarity, Hamming distance, Mahalanobis distance, correlation coefficient, and KL divergence. In the store matching method provided in the embodiment of the present application, the preset store that matches the question information is determined by combining direct matching rules with semantic matching rules, which can improve the accuracy of store matching.

[0090] See also Figure 7 , Figure 72 is a schematic diagram of the structure of the store matching device provided in the embodiment of the present application. In some embodiments, the store matching device 20 may include a plurality of functional modules composed of computer program segments. The computer programs of the various program segments in the store matching device 20 may be stored in the memory of the computer device 10 and executed by at least one controller to execute (see Figure 2 Description) Store matching function.

[0091] In this embodiment, the store matching device 20 can be divided into multiple functional modules according to the functions it performs. The functional modules may include: a question receiving module 201, a candidate determination module 202, a keyword combination module 203, and a store matching module 204. The module referred to in this application refers to a series of computer program segments that can be executed by at least one controller and can complete fixed functions, which are stored in a memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0092] The question receiving module 201 may be used to receive question information, parse the question information, and obtain corresponding initial keywords.

[0093] The candidate determination module 202 may be configured to determine candidate keywords whose semantic similarity with the initial keyword is greater than or equal to a first threshold.

[0094] The keyword combination module 203 can be used to combine the initial keyword with the candidate keyword to obtain a target keyword.

[0095] The store matching module 204 may be used to determine a preset store that matches the question information based on a preset store tag and the target keyword.

[0096] It can be understood that the store matching device 20 and the store matching method of the above embodiment belong to the same inventive concept, and the specific implementation method of each module of the store matching device 20 corresponds to the steps of store matching in the above embodiment, which will not be repeated in this application.

[0097] The module division described above is a logical function division, and there may be other division methods in actual implementation. In addition, the functional modules in each embodiment of the present application may be integrated in the same processing unit, or each module may exist physically separately, or two or more modules may be integrated in the same unit. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0098] then Figure 1The description of the computer device is that a computer program is stored in the memory 11, and when the computer program is executed by at least one controller 12, all or part of the steps in the store matching are implemented. The memory 11 includes a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0099] Furthermore, the computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the computer device 10, etc.

[0100] In some embodiments, at least one controller 12 is the control core (ControlUnit) of the computer device 10, and uses various interfaces and lines to connect various components of the entire computer device 10, and executes various functions and processes data of the computer device 10 by running or executing programs or modules stored in the memory 11, and calling data stored in the memory 11. For example, when at least one controller 12 executes the computer program stored in the memory, it implements all or part of the steps of the store matching in the embodiment of the present application; or implements all or part of the functions of the store matching device. At least one controller 12 can be composed of an integrated circuit, for example, it can be composed of a single packaged integrated circuit, or it can be composed of multiple integrated circuits with the same function or different functions, including one or more central processing units (CPUs), microcontrollers, digital processing chips, graphics processors, and combinations of various control chips.

[0101] The above-mentioned integrated unit implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a computer device, or a network device, etc.) or a controller (processor) to execute parts of the methods of various embodiments of the present application.

[0102] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules is only a logical function division, and there may be other division methods in actual implementation.

[0103] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0104] In addition, each functional module in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0105] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the spirit or basic features of the present application. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present application is limited by the attached claims rather than the above description, so it is intended to include all changes that fall within the meaning and scope of the equivalent elements of the claims in the present application. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "including" does not exclude other units or, and the singular does not exclude the plural. Multiple units or devices stated in the specification can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical solution of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present application.

Claims

1. A store matching method, characterized in that: The store matching method comprises: Receiving question information, parsing the question information, and obtaining corresponding initial keywords; Determine candidate keywords whose semantic similarity with the initial keyword is greater than or equal to a first threshold; Combining the initial keyword with the candidate keyword to obtain a target keyword; A preset store matching the question information is determined based on the preset store tag and the target keyword.

2. The store matching method according to claim 1, characterized in that: Before receiving the question information, parsing the question information, and obtaining the corresponding initial keywords, the method further includes: Determine a question description corresponding to the question information; Determine the question category corresponding to the question information according to the corresponding relationship between the question description and the question category; If the question is classified as a target category, the question information is parsed to obtain the initial keywords.

3. The store matching method according to claim 1, characterized in that: The step of parsing the question information to obtain corresponding initial keywords includes: Performing word segmentation processing on the question information to obtain multiple word segments; Determine the segmentation attribute corresponding to each segmentation; The segmented word corresponding to the target segmented word attribute is used as the initial keyword.

4. The store matching method according to claim 1, characterized in that: The step of determining a candidate keyword having a semantic similarity with the initial keyword greater than or equal to a first threshold comprises: Based on the preset corresponding relationship between the initial keyword and the candidate keyword, a candidate keyword having a semantic similarity with the initial keyword greater than or equal to a first threshold is determined.

5. The store matching method according to claim 1, characterized in that: The method further comprises: Determine the business categories and merchandise corresponding to the preset store; Determine a first label of the preset store according to the business category, and determine a second label of the preset store according to the business commodity; The first tag and the second tag are combined to obtain the store tag.

6. The store matching method according to claim 1, characterized in that: The step of determining a preset store matching the question information based on the preset store tag and the target keyword includes: Determine the preset store corresponding to the store tag that is the same as any target keyword; Determine the number of stores of the matching preset stores; If the number of stores is less than a second threshold, semantic matching is performed based on the store tags and the target keywords to obtain a semantic matching result; A preset store matching the question information is determined according to the semantic matching result.

7. The store matching method according to claim 6, characterized in that: The performing semantic matching based on the store tag and the target keyword to obtain a semantic matching result includes: Determine a first vector corresponding to the store label; Determine a second vector corresponding to the target keyword; If the similarity between the first vector and the second vector is greater than a third threshold, determining that the semantic matching result is a semantic match between the store tag and the target keyword; If the similarity between the first vector and the second vector is less than or equal to the third threshold, it is determined that the semantic matching result is that the store tag does not semantically match the target keyword.

8. The store matching method according to claim 6, characterized in that: The method further comprises: If the number of stores is greater than or equal to the second threshold, sorting the matching preset stores according to a preset rule; A preset number of preset stores ranked first are selected as preset stores matching the question information.

9. A store matching device, characterized in that: The store matching device comprises: A question receiving module, used to receive question information, analyze the question information, and obtain corresponding initial keywords; A candidate determination module, used to determine candidate keywords whose semantic similarity with the initial keyword is greater than or equal to a first threshold; A keyword combination module, used for combining the initial keyword with the candidate keyword to obtain a target keyword; The store matching module is used to determine a preset store that matches the question information based on a preset store tag and the target keyword.

10. A computer device, characterized in that: The computer device includes a controller and a memory, and the controller is used to implement the store matching method according to any one of claims 1 to 8 when executing the computer program stored in the memory.

Images