Item recommendation method, device, electronic device and storage medium

Abstract

The present invention relates to an item recommendation method, device, electronic device, and storage medium. The item recommendation method includes: obtaining keywords input by a user regarding an item to be recommended and personalized information of the user; determining a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for multiple items; determining label words and / or label word combinations that match the keywords in the second knowledge graph; determining items that match the label words and / or label word combinations based on preset item-label correspondences, and recommending them. The embodiment of the present application automatically recommends items or item combinations that meet the user's personalized needs based on the keywords input by the user without the user needing to understand the parameters of the items, thereby improving the efficiency of the user's item selection and facilitating user use.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to an item recommendation method, device, electronic device, and storage medium. Background Art

[0002] At present, various home appliances have more and more functions, and different house types and rooms of different sizes require different home appliances.

[0003] However, most users do not have sufficient knowledge of the product names and parameters, which results in a lot of homework when choosing home appliances for their families, taking up a lot of time and being very inefficient. Summary of the Invention

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the present application provides an item recommendation method, device, electronic device and storage medium.

[0005] In a first aspect, the present application provides an item recommendation method, comprising:

[0006] Obtaining keywords input by the user regarding the item to be recommended and the user's personalized information;

[0007] Determining a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for a plurality of items;

[0008] Determining, in the second knowledge graph, a tag word and / or a tag word combination that matches the keyword;

[0009] Based on the preset item label correspondence, items that match the label words and / or label word combinations are determined and recommended.

[0010] Optionally, determining the second knowledge graph based on the personalized information and a preset first knowledge graph includes:

[0011] Extracting environmental words from the personalized information, where the environmental words are used to indicate environmental information of the usage environment of the item to be recommended;

[0012] Converting the environment word into a first label word;

[0013] Determine in the first knowledge graph a label word that matches the first label word to obtain a second knowledge graph.

[0014] Optionally, determining the second knowledge graph based on the personalized information and a preset first knowledge graph includes:

[0015] Extracting attribute words from the personalized information, wherein the attribute words are used to indicate attribute information of the item to be recommended;

[0016] Obtaining a second label word corresponding to the attribute word;

[0017] Determine in the first knowledge graph a label word that matches the second label word to obtain a second knowledge graph.

[0018] Optionally, determining a tag word and / or tag word combination matching the keyword in the second knowledge graph includes:

[0019] If the number of the keywords is at least two, determining whether each of the keywords contains an item identifier;

[0020] If each of the keywords contains an item identifier, determining a tag category corresponding to the item identifier in the second knowledge graph, and matching the keywords other than the item identifier in at least two keywords with each tag word under the tag category to obtain a tag word and / or tag word combination that matches the keyword;

[0021] Alternatively, if the keyword does not contain an item identifier, determine whether each of the keywords has a unique corresponding item identifier; if any of the keywords has a unique corresponding item identifier, determine the label category corresponding to the item identifier in the second knowledge graph, and match the keywords other than the item identifier in at least two keywords with each label word under the label category to obtain label words and / or label word combinations that match the keywords.

[0022] Optionally, constructing the first knowledge graph includes:

[0023] Obtain identification information, attribute information and evaluation information of multiple items;

[0024] constructing an item information table based on the identification information, the attribute information, and the evaluation information of the plurality of items, wherein the item information table includes a plurality of label words;

[0025] Obtaining derivative words corresponding to each tag word in the item information table;

[0026] The first knowledge graph is constructed based on the plurality of label words and the derivative words.

[0027] Optionally, constructing an item information table based on the identification information, the attribute information, and the evaluation information of a plurality of items includes:

[0028] For each item, convert the identification information and the attribute information into label words respectively;

[0029] Segmenting the evaluation information and determining whether there are new words in the segmented words;

[0030] If there is no new word in the words obtained by the word segmentation, the words obtained by the word segmentation are determined as label words;

[0031] If there are new words in the evaluation information, classify the new words to obtain new word categories; if the new words are words related to items, determine the new words as label words;

[0032] Build an item information table based on each tag word.

[0033] Optionally, obtaining derivative words corresponding to each tag word in the item information table includes:

[0034] Obtaining synonyms corresponding to each tag word in the item information table;

[0035] Obtaining spoken words corresponding to each tag word in the item information table;

[0036] The synonyms and the spoken words are determined as derivative words corresponding to the tag word.

[0037] In a second aspect, the present application provides an item recommendation device, comprising:

[0038] A first acquisition module is used to acquire keywords input by a user regarding items to be recommended and personalized information of the user;

[0039] A first determining module is configured to determine a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for a plurality of items;

[0040] A second determination module is used to determine a label word and / or a label word combination that matches the keyword in the second knowledge graph;

[0041] The third determining module is used to determine items that match the label words and / or label word combinations based on preset item label correspondences and recommend them.

[0042] Optionally, the first determining module includes:

[0043] a first extraction unit, configured to extract environmental words from the personalized information, wherein the environmental words are used to indicate environmental information of a usage environment of the item to be recommended;

[0044] a conversion unit, configured to convert the environment word into a first label word;

[0045] The first determination unit is used to determine a label word that matches the first label word in the first knowledge graph to obtain a second knowledge graph.

[0046] Optionally, the first determining module includes:

[0047] a second extraction unit, configured to extract attribute words from the personalized information, wherein the attribute words are used to indicate attribute information of the item to be recommended;

[0048] A first acquiring unit, configured to acquire a second label word corresponding to the attribute word;

[0049] The second determination unit is used to determine the label word that matches the second label word in the first knowledge graph to obtain a second knowledge graph.

[0050] Optionally, the second determining module includes:

[0051] a third determining unit, configured to determine whether each of the keywords contains an item identifier if the number of the keywords is at least two;

[0052] a fourth determining unit, configured to, if each of the keywords includes an item identifier, determine a tag category corresponding to the item identifier in the second knowledge graph, and match keywords other than the item identifier among at least two keywords with each tag word under the tag category to obtain a tag word and / or tag word combination that matches the keyword;

[0053] Alternatively, the fifth determination unit is used to determine whether each of the keywords has a unique corresponding item identifier if the keyword does not contain an item identifier; if any of the keywords has a unique corresponding item identifier, determine the label category corresponding to the item identifier in the second knowledge graph; and match the keywords other than the item identifier in at least two keywords with each label word under the label category to obtain label words and / or label word combinations that match the keywords.

[0054] Optionally, the device further comprises:

[0055] A second acquisition module is used to obtain identification information, attribute information and evaluation information of multiple items;

[0056] A first constructing module is configured to construct an item information table based on the identification information, the attribute information, and the evaluation information of a plurality of items, wherein the item information table includes a plurality of tag words;

[0057] A third acquisition module is used to acquire derivative words corresponding to each tag word in the item information table;

[0058] The second construction module is used to construct the first knowledge graph based on multiple label words and the derivative words.

[0059] Optionally, the first building block includes:

[0060] a conversion unit, configured to convert the identification information and the attribute information into label words for each item;

[0061] a sixth determining unit, configured to segment the evaluation information and determine whether there are new words in the segmented words;

[0062] a seventh determining unit, configured to determine the words obtained by word segmentation as label words if no new words exist in the words obtained by word segmentation;

[0063] an eighth determining unit, configured to classify a new word in the evaluation information to obtain a new word category if the new word exists in the evaluation information, and determine the new word as a tag word if the new word is an item-related word;

[0064] The construction unit is used to construct an item information table based on each tag word.

[0065] Optionally, the third acquisition module includes:

[0066] A second acquiring unit is configured to acquire synonyms corresponding to each tag word in the item information table;

[0067] a third acquiring unit, configured to acquire spoken words corresponding to each label word in the item information table;

[0068] The ninth determining unit is configured to determine the synonyms and the spoken words as derivative words corresponding to the tag word.

[0069] In a third aspect, the present application provides an electronic device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0070] Memory for storing computer programs;

[0071] The processor is configured to implement any item recommendation method described in the first aspect when executing a program stored in the memory.

[0072] In a fourth aspect, the present application provides a computer-readable storage medium, on which a program of an item recommendation method is stored. When the program of the item recommendation method is executed by a processor, the steps of any one of the item recommendation methods described in the first aspect are implemented.

[0073] The above technical solution provided by the embodiment of the present application has the following advantages compared with the prior art:

[0074] In an embodiment of the present application, by automatically determining a second knowledge graph corresponding to the user's personalized information in the first knowledge graph, and determining label words and / or label word combinations based on the keywords input by the user in the second knowledge graph, and then determining items to be recommended based on the label words and / or label word combinations, and recommending items or item combinations, the user does not need to understand the parameters of the items, and the items or item combinations that meet the user's personalized needs can be automatically recommended to the user based on the keywords input by the user, thereby improving the efficiency of user item selection and facilitating user use. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0076] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0077] Figure 1 A flowchart of an item recommendation method provided in an embodiment of the present application;

[0078] Figure 2 A structural diagram of an item recommendation device provided in an embodiment of the present application;

[0079] Figure 3 A structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0080] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0081] Since most users currently do not have sufficient knowledge of product names and parameters, they need to do a lot of research when selecting home appliances for their families, which takes up a lot of time and is very inefficient. To this end, the embodiments of the present application provide an item recommendation method, device, electronic device, and storage medium.

[0082] The item recommendation method provided in the embodiment of the present application can be applied to a terminal device. For example, the terminal device can be a server, a client, etc. Figure 1This is a flowchart of an item recommendation method provided in an embodiment of the present application, such as Figure 1 As shown, the item recommendation method may include the following steps:

[0083] Step S101, obtaining keywords related to items to be recommended and personalized information of the user inputted by the user;

[0084] When a user needs to recommend an item, he or she can enter the keyword of the item to be recommended, that is, the keyword of the item to be recommended. For example, if the item the user needs to recommend is an air conditioner, he or she can enter air conditioner, etc. The user's personalized information is pre-configured by the user. For example, a small amount of information can be collected during the user registration stage. The personalized information is used to indicate the user's personalized preferences for performance, appearance, usage location, etc. For example, the personalized information may include: house type, area, room type, decoration style, special needs, color, energy saving, quietness, living habits, overall budget, etc. For example, some home appliances with labels such as air circulation, purifier, and quiet can be recommended in some bedrooms and office rooms. Some rooms have lighting, smart control and other requirements, which also need to be recommended.

[0085] By obtaining personalized information, it is possible to provide users with personalized customized home appliance recommendations based on the size of their room, their needs for special home appliance functions, their living habits, and their overall budget.

[0086] Step S102: determining a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for a plurality of items;

[0087] In this step, matching label words can be filtered in the first knowledge graph based on the personalized information, and then the filtered label words can be used to determine the second knowledge graph.

[0088] Step S103: determining a tag word and / or tag word combination matching the keyword in the second knowledge graph;

[0089] In this step, the keywords input by the user can be matched with the label words in the second knowledge graph to obtain label words and / or label word combinations that match the keywords, wherein the label word combination includes at least two label words corresponding to different items.

[0090] Step S104: determining items that match the tag word and / or tag word combination based on the preset item-tag correspondence, and recommending them.

[0091] In an embodiment of the present application, an item label correspondence relationship can be pre-set, that is, the correspondence relationship between an item and a label word, so that any label word can be mapped to the product. After obtaining the label word and / or label word combination that matches the keyword in step S103, the item label correspondence relationship can be searched for items that match the label word, and / or items that match each label word in the label word combination, and the found items and / or item combinations can be recommended.

[0092] In an embodiment of the present application, by automatically determining a second knowledge graph corresponding to the user's personalized information in the first knowledge graph, and determining label words and / or label word combinations based on the keywords input by the user in the second knowledge graph, and then determining items to be recommended based on the label words and / or label word combinations, and recommending items or item combinations, the user does not need to understand the parameters of the items, and the items or item combinations that meet the user's personalized needs can be automatically recommended to the user based on the keywords input by the user, thereby improving the efficiency of user item selection and facilitating user use.

[0093] In another embodiment of the present application, step S102 determines the second knowledge graph based on the personalized information and the preset first knowledge graph, including:

[0094] Step S201, extracting environmental words from the personalized information;

[0095] In an embodiment of the present application, the environmental words are used to indicate environmental information of the usage environment of the recommended item, that is, the environmental words can reflect the usage environment of the item, such as: location, house type, area, decoration style, etc.

[0096] Step S202, converting the environmental word into a first label word;

[0097] Since there may be no environment-related words in the first knowledge graph, the environment words can be converted. Furthermore, the relevant house types can be automatically searched from the Internet based on the location, house type and area in the environment words, and the area of ​​the specified rooms in the house type (such as living room and bedrooms) can be determined, and then the first label word (such as 1.5 horsepower) can be determined based on the area of ​​the specified room; the color matching can also be automatically searched from the Internet based on the decoration style in the environment words, and the first label word (such as white) can be determined based on the color matching.

[0098] Step S203: Determine the label words that match the first label words in the first knowledge graph to obtain a second knowledge graph.

[0099] The embodiment of the present application extracts environmental words and converts them into first label words, and then determines the second knowledge graph based on the first label words, thereby filtering the first knowledge graph based on the environmental information in the user's personalized information to obtain the second knowledge graph, understand the user's personalized needs, and facilitate automatic recommendation to the user of items or item combinations that meet the user's personalized needs.

[0100] In another embodiment of the present application, step S102 determines the second knowledge graph based on the personalized information and the preset first knowledge graph, including:

[0101] Step S301, extracting attribute words from the personalized information;

[0102] In an embodiment of the present application, the attribute words are used to indicate the attribute information of the item to be recommended, that is, the attribute words can reflect the attributes of the item, such as: color, size, energy saving, noise level, price, power consumption, additional functions, etc.

[0103] Step S302: obtaining a second label word corresponding to the attribute word;

[0104] The personalized information input by the user may contain colloquial attribute words or attribute words in non-standard languages. Therefore, the second label word corresponding to the threshold can be obtained based on the attribute word, which facilitates matching with the label word in the first knowledge graph.

[0105] Step S303: Determine the label word that matches the second label word in the first knowledge graph to obtain a second knowledge graph.

[0106] The embodiment of the present application extracts attribute words and obtains second label words corresponding to the attribute words, and then determines the second knowledge graph based on the second label words, thereby filtering the first knowledge graph based on the attribute information of the items in the user's personalized information to obtain the second knowledge graph, understand the user's personalized needs, and facilitate automatic recommendation to the user of items or item combinations that meet the user's personalized needs.

[0107] In another embodiment of the present application, determining a tag word and / or tag word combination matching the keyword in the second knowledge graph includes:

[0108] If the number of keywords is one, the keyword can be directly matched with each label word in the second knowledge graph. If the match is successful, a label word matching the keyword is obtained.

[0109] If the number of the keywords is at least two, determine whether each of the keywords contains an item identifier; in an embodiment of the present application, the item identifier can be used to uniquely identify an item. For example, the item identifier can refer to the item name, item model or item ID, etc.

[0110] If each of the keywords contains an item identifier, determining a tag category corresponding to the item identifier in the second knowledge graph, and matching the keywords other than the item identifier in at least two keywords with each tag word under the tag category to obtain a tag word and / or tag word combination that matches the keyword;

[0111] In the second knowledge graph, each item identifier can be used as a basis for classifying label words, so the label category can be determined based on the item identifier first, and then under the label category, label words and / or label word combinations that match the keywords (except the item identifier) ​​can be searched.

[0112] Alternatively, if the keyword does not contain an item identifier, determine whether each of the keywords has a unique corresponding item identifier; if any of the keywords has a unique corresponding item identifier, determine the label category corresponding to the item identifier in the second knowledge graph, and match the keywords other than the item identifier in at least two keywords with each label word under the label category to obtain label words and / or label word combinations that match the keywords.

[0113] If the keyword does not contain the item identifier, it can be determined whether an item identifier can be uniquely determined based on each keyword. If so, the tag category can be determined first based on the item identifier, and then under this tag category, tag words and / or tag word combinations that match the keyword (except the item identifier) ​​can be searched.

[0114] The embodiment of the present application can automatically determine the corresponding label words in the second knowledge graph based on the keywords, further filter the label words based on the keywords, and then automatically recommend items or item combinations that meet the user's personalized needs to the user based on the keywords entered by the user, thereby improving the efficiency of user item selection and facilitating user use.

[0115] In another embodiment of the present application, constructing the first knowledge graph includes:

[0116] Step S401, obtaining identification information, attribute information and evaluation information of multiple items;

[0117] By mining data sources such as item libraries and e-commerce platform data, basic product information can be obtained, such as identification information and attribute information (such as product parameters, appearance and other basic data).

[0118] Step S402: constructing an item information table based on the identification information, the attribute information, and the evaluation information of the plurality of items, wherein the item information table includes a plurality of tag words;

[0119] The identification information and attribute information are stored in the item information table. In order to be able to make recommendations from more aspects in the subsequent recommendation process, such as whether the product can be connected to the Internet, whether it can be voice-operated, whether it is quiet enough, and other non-basic parameter information, the identification information and the attribute information can be further expanded, that is, converted into label words.

[0120] In one embodiment of the present application, step S402 constructs an item information table based on the identification information, the attribute information, and the evaluation information of the plurality of items, including:

[0121] Step S501: for each item, convert the identification information and the attribute information into label words;

[0122] Step S502: segmenting the evaluation information and determining whether there are new words in the segmented words;

[0123] The embodiment of the present application can also continue to mine online product reviews, store reviews and other data, and continue to perform word segmentation and new word discovery on the mined words and sentences related to each product. Word segmentation is to split the mined words and sentences, split the text into sentences, and divide the sentences into words. After word segmentation is completed, words that are not in the current product information table can be marked as new words.

[0124] Step S503: If there is no new word in the words obtained by word segmentation, the words obtained by word segmentation are determined as label words;

[0125] In step S504, if new words are present in the review information, they are classified to determine their categories. If the new words are item-related, they are identified as label words. A binary classification model is trained using a small number of samples (to classify the new words into useless words and words related to product information, and then iterate on the words related to product information from the previous round). The new words are then classified using the binary classification model. Furthermore, through the integration of machine learning, words that are difficult to distinguish from the previous round of results are selected and re-labeled, and the classification iteration continues until convergence.

[0126] Step S505: construct an item information table based on each tag word.

[0127] Step S403, obtaining derivative words corresponding to each tag word in the item information table;

[0128] In this step, the label words in the item information table can be counted, and many derivative words can be mapped for each type of information.

[0129] In one embodiment of the present application, step S403 obtains derivative words corresponding to each tag word in the item information table, including:

[0130] Obtain synonyms corresponding to each label word in the item information table; obtain spoken words corresponding to each label word in the item information table; and determine the synonyms and the spoken words as derivative words corresponding to the label words.

[0131] Label words can be synonyms for the product information or colloquial terms. For example, a 1.5P air conditioner could have labels like "1.5 HP" and "suitable room size range." A quiet product could have synonyms like "quiet," "silent," and "noiseless." The classification model described above can be used to iterate and identify related words such as synonyms and colloquial terms as label words for the information.

[0132] Step S404: construct the first knowledge graph based on the multiple label words and the derivative words.

[0133] The embodiment of the present application can automatically construct a first knowledge graph, and then facilitate automatically determining a second knowledge graph corresponding to the user's personalized information in the first knowledge graph, and determine label words and / or label word combinations in the second knowledge graph based on the keywords input by the user, and then determine the items to be recommended based on the label words and / or label word combinations, and recommend items or item combinations. Without the user knowing the parameters of the items, the user can be automatically recommended items or item combinations that meet the user's personalized needs based on the keywords input by the user, thereby improving the efficiency of user item selection and facilitating user use.

[0134] In another embodiment of the present application, an item recommendation device is also provided. Figure 2 Shown, including:

[0135] A first acquisition module 11 is configured to acquire keywords input by a user regarding items to be recommended and personalized information of the user;

[0136] A first determining module 12 is configured to determine a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for a plurality of items;

[0137] A second determining module 13 is configured to determine, in the second knowledge graph, a tag word and / or a tag word combination that matches the keyword;

[0138] The third determining module 14 is configured to determine and recommend items that match the tag word and / or tag word combination based on a preset item-tag correspondence relationship.

[0139] Optionally, the first determining module includes:

[0140] a first extraction unit, configured to extract environmental words from the personalized information, wherein the environmental words are used to indicate environmental information of a usage environment of the item to be recommended;

[0141] a conversion unit, configured to convert the environment word into a first label word;

[0142] The first determination unit is used to determine a label word that matches the first label word in the first knowledge graph to obtain a second knowledge graph.

[0143] Optionally, the first determining module includes:

[0144] a second extraction unit, configured to extract attribute words from the personalized information, wherein the attribute words are used to indicate attribute information of the item to be recommended;

[0145] A first acquiring unit, configured to acquire a second label word corresponding to the attribute word;

[0146] The second determination unit is used to determine the label word that matches the second label word in the first knowledge graph to obtain a second knowledge graph.

[0147] Optionally, the second determining module includes:

[0148] a third determining unit, configured to determine whether each of the keywords contains an item identifier if the number of the keywords is at least two;

[0149] a fourth determining unit, configured to, if each of the keywords includes an item identifier, determine a tag category corresponding to the item identifier in the second knowledge graph, and match keywords other than the item identifier among at least two keywords with each tag word under the tag category to obtain a tag word and / or tag word combination that matches the keyword;

[0150] Alternatively, the fifth determination unit is used to determine whether each of the keywords has a unique corresponding item identifier if the keyword does not contain an item identifier; if any of the keywords has a unique corresponding item identifier, determine the label category corresponding to the item identifier in the second knowledge graph; and match the keywords other than the item identifier in at least two keywords with each label word under the label category to obtain label words and / or label word combinations that match the keywords.

[0151] Optionally, the device further comprises:

[0152] A second acquisition module is used to obtain identification information, attribute information and evaluation information of multiple items;

[0153] A first constructing module is configured to construct an item information table based on the identification information, the attribute information, and the evaluation information of a plurality of items, wherein the item information table includes a plurality of tag words;

[0154] A third acquisition module is used to acquire derivative words corresponding to each tag word in the item information table;

[0155] The second construction module is used to construct the first knowledge graph based on multiple label words and the derivative words.

[0156] Optionally, the first building block includes:

[0157] a conversion unit, configured to convert the identification information and the attribute information into label words for each item;

[0158] a sixth determining unit, configured to segment the evaluation information and determine whether there are new words in the segmented words;

[0159] a seventh determining unit, configured to determine the words obtained by word segmentation as label words if no new words exist in the words obtained by word segmentation;

[0160] an eighth determining unit, configured to classify a new word in the evaluation information to obtain a new word category if the new word exists in the evaluation information, and determine the new word as a tag word if the new word is an item-related word;

[0161] The construction unit is used to construct an item information table based on each tag word.

[0162] Optionally, the third acquisition module includes:

[0163] A second acquiring unit is configured to acquire synonyms corresponding to each tag word in the item information table;

[0164] a third acquiring unit, configured to acquire spoken words corresponding to each label word in the item information table;

[0165] The ninth determining unit is configured to determine the synonyms and the spoken words as derivative words corresponding to the tag word.

[0166] In another embodiment of the present application, an electronic device is provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0167] Memory for storing computer programs;

[0168] The processor is configured to implement the item recommendation method described in any of the aforementioned method embodiments when executing the program stored in the memory.

[0169] In the electronic device provided by an embodiment of the present invention, the processor realizes by executing a program stored in the memory, automatically determining a second knowledge graph corresponding to the user's personalized information in the first knowledge graph, and determining label words and / or label word combinations based on the keywords input by the user in the second knowledge graph, and then determining items to be recommended based on the label words and / or label word combinations, and recommending items or item combinations. Without the user knowing the parameters of the items, the processor can automatically recommend items or item combinations that meet the user's personalized needs based on the keywords input by the user, thereby improving the efficiency of the user's item selection and facilitating user use.

[0170] The communication bus 1140 mentioned in the above electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The communication bus 1140 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0171] The communication interface 1120 is used for communication between the electronic device and other devices.

[0172] The memory 1130 may include a random access memory (RAM) or a non-volatile memory, such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0173] The above-mentioned processor 1110 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0174] In another embodiment of the present application, a computer-readable storage medium is provided, on which a program of an item recommendation method is stored. When the program of the item recommendation method is executed by a processor, the steps of the item recommendation method described in any of the aforementioned method embodiments are implemented.

[0175] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0176] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is intended to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An item recommendation method, characterized in that: include: Obtaining keywords input by the user regarding the item to be recommended and the user's personalized information; Determining a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for a plurality of items; Determining a second knowledge graph based on the personalized information and a preset first knowledge graph includes: Extracting environmental words from the personalized information, where the environmental words are used to indicate environmental information of the usage environment of the item to be recommended; Converting the environmental word into a first label word includes: searching for relevant apartment types online based on the location, apartment type, and area in the environmental word, determining the area of ​​a specified room in the apartment type, and then determining the first label word based on the area of ​​the specified room; or searching for color combinations online based on the decoration style in the environmental word, and determining the first label word based on the color combinations; Determining, in the first knowledge graph, a label word that matches the first label word to obtain a second knowledge graph; Determining, in the second knowledge graph, a tag word and / or a tag word combination that matches the keyword; Based on the preset item label correspondence, items that match the label words and / or label word combinations are determined and recommended.

2. The method according to claim 1, characterized in that Determining a second knowledge graph based on the personalized information and a preset first knowledge graph includes: Extracting attribute words from the personalized information, wherein the attribute words are used to indicate attribute information of the item to be recommended; Obtaining a second label word corresponding to the attribute word; Determine in the first knowledge graph a label word that matches the second label word to obtain a second knowledge graph.

3. The method according to claim 1, characterized in that Determining a tag word and / or tag word combination matching the keyword in the second knowledge graph includes: If the number of the keywords is at least two, determining whether each of the keywords contains an item identifier; If each of the keywords contains an item identifier, determining a tag category corresponding to the item identifier in the second knowledge graph, and matching the keywords other than the item identifier in at least two keywords with each tag word under the tag category to obtain a tag word and / or tag word combination that matches the keyword; Alternatively, if the keyword does not contain an item identifier, determine whether each of the keywords has a unique corresponding item identifier; if any of the keywords has a unique corresponding item identifier, determine the label category corresponding to the item identifier in the second knowledge graph, and match the keywords other than the item identifier in at least two keywords with each label word under the label category to obtain label words and / or label word combinations that match the keywords.

4. The method according to claim 1, wherein Constructing the first knowledge graph includes: Obtain identification information, attribute information and evaluation information of multiple items; constructing an item information table based on the identification information, the attribute information, and the evaluation information of the plurality of items, wherein the item information table includes a plurality of label words; Obtaining derivative words corresponding to each tag word in the item information table; The first knowledge graph is constructed based on the plurality of label words and the derivative words.

5. The method according to claim 4, characterized in that Constructing an item information table based on the identification information, the attribute information, and the evaluation information of a plurality of items includes: For each item, convert the identification information and the attribute information into label words respectively; Segmenting the evaluation information and determining whether there are new words in the segmented words; If there is no new word in the words obtained by the word segmentation, the words obtained by the word segmentation are determined as label words; If there are new words in the evaluation information, classify the new words to obtain new word categories; if the new words are words related to items, determine the new words as label words; Build an item information table based on each tag word.

6. The method according to claim 4, characterized in that Obtaining derivative words corresponding to each tag word in the item information table includes: Obtaining synonyms corresponding to each tag word in the item information table; Obtaining spoken words corresponding to each tag word in the item information table; The synonyms and the spoken words are determined as derivative words corresponding to the tag word.

7. An item recommendation device, characterized in that: include: A first acquisition module is used to acquire keywords input by a user regarding items to be recommended and personalized information of the user; A first determining module is configured to determine a second knowledge graph based on the personalized information and a preset first knowledge graph, wherein the first knowledge graph includes label words for a plurality of items; Determining a second knowledge graph based on the personalized information and a preset first knowledge graph includes: extracting environmental words from the personalized information, the environmental words being used to indicate environmental information of the usage environment of the item to be recommended; converting the environmental words into first label words, including: searching for relevant apartment types from the Internet based on the location, apartment type, and area in the environmental words, determining the area of ​​a specified room in the apartment type, and then determining the first label word based on the area of ​​the specified room; or searching for color combinations from the Internet based on the decoration style in the environmental words, and determining the first label word based on the color combinations; and determining a label word in the first knowledge graph that matches the first label word to obtain a second knowledge graph; A second determination module is used to determine a label word and / or a label word combination that matches the keyword in the second knowledge graph; The third determining module is used to determine items that match the label words and / or label word combinations based on preset item label correspondences and recommend them.

8. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; A processor, configured to implement the item recommendation method according to any one of claims 1 to 6 when executing a program stored in a memory.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program of the item recommendation method, and when the program of the item recommendation method is executed by a processor, the steps of the item recommendation method according to any one of claims 1 to 6 are implemented.

Images