WO2020215751A1 - 一种推荐方法、电子设备及可读存储介质 - Google Patents

一种推荐方法、电子设备及可读存储介质 Download PDF

Info

Publication number
WO2020215751A1
WO2020215751A1 PCT/CN2019/125713 CN2019125713W WO2020215751A1 WO 2020215751 A1 WO2020215751 A1 WO 2020215751A1 CN 2019125713 W CN2019125713 W CN 2019125713W WO 2020215751 A1 WO2020215751 A1 WO 2020215751A1
Authority
WO
WIPO (PCT)
Prior art keywords
historical
commodity
information
keywords
user
Prior art date
Application number
PCT/CN2019/125713
Other languages
English (en)
French (fr)
Inventor
史亦隆
曾轲
李容
Original Assignee
北京三快在线科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京三快在线科技有限公司 filed Critical 北京三快在线科技有限公司
Publication of WO2020215751A1 publication Critical patent/WO2020215751A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce
    • G06Q30/06Buying, selling or leasing transactions
    • G06Q30/0601Electronic shopping [e-shopping]
    • G06Q30/0631Item recommendations

Definitions

  • the present disclosure relates to the technical field of recommendation, in particular to a recommendation method, a recommendation device, electronic equipment, and a readable storage medium
  • Obtaining the user's consumption tendency according to the search behavior information provided by the user and recommending the corresponding product for the user is a widely used commodity recommendation method at present.
  • the recommendation is generally based on the user's historical behavior data, product popularity and other information.
  • the disadvantage is that the generalization performance of the product is relatively poor, and the user can only recommend the user to browse or Products with relatively high popularity, and current users often give out a pan-concept product (the so-called pan-concept product generally refers to a certain category of goods, the category can be large or small, such as "a bag of rice", "a bag of 10KG rice”, “A bag of 10KGXX rice” is three categories from big to small), not a product with very clear demand; on the other hand, users are more keen on those products with high popularity, large discount rate and high cost performance, but currently Merchants have various promotion methods, and it is difficult for users to choose a truly affordable and popular product from different merchants.
  • the embodiments of the present disclosure are proposed to provide a recommendation method and a corresponding recommendation device that overcome the above problems or at least partially solve the above problems.
  • the embodiments of the present disclosure disclose a recommendation method, which specifically includes: acquiring historical information of a user; determining the key needs of the user according to the historical information; extracting from a preset merchant list database Candidate products that match the key needs; extract the product sales information of the candidate products; obtain, according to the product sales information, a product sales cost-effective index that meets the key needs; according to the product sales cost-effective index, provide to the user Display products for the product sales information.
  • an embodiment of the present disclosure discloses a recommendation device, which specifically includes: a historical information acquisition module for acquiring historical information of a user; a key requirement determination module for determining the historical information according to the historical information. The user’s key needs; a candidate product extraction module for extracting candidate products matching the key needs in a preset merchant list database; a product sales information extraction module for extracting product sales information of the candidate product; cost-effective index
  • the determining module is used to obtain the commodity sales price/performance index that meets the key requirements according to the commodity sales information; the commodity display module is used to display to the user information specific to the commodity sales information according to the commodity sales price/performance index commodity.
  • an electronic device including: a processor, a memory, and a computer program stored on the memory and running on the processor, wherein the processor executes The program implements the aforementioned recommended method.
  • a readable storage medium which when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to implement the aforementioned recommended method.
  • the embodiments of the present disclosure include the following advantages: by obtaining the user's historical information, the user's general needs, that is, the key needs, are found, and the sellers of products corresponding to the key needs are found, and products with higher cost performance are recommended to the users. It has the ability to effectively understand the general needs in the user's historical information, provide users with the most cost-effective products according to the general needs of users, and reduce the time-consuming beneficial effect of further screening and comparison of similar products by users.
  • FIG. 1 is a step flowchart of an embodiment of a recommendation method according to an embodiment of the present disclosure
  • FIG. 2 is a step flowchart of an embodiment of a recommendation method according to an embodiment of the present disclosure
  • FIG. 3 is a structural block diagram of an embodiment of a recommendation device according to an embodiment of the present disclosure
  • FIG. 4 is a structural block diagram of an embodiment of a recommendation device according to an embodiment of the present disclosure.
  • Fig. 5 schematically shows a block diagram of a computing processing device for executing a method according to an embodiment of the present disclosure
  • Fig. 6 schematically shows a storage unit for holding or carrying program codes for implementing the method according to the embodiments of the present disclosure.
  • Step 101 Obtain user history information
  • recommendations are usually made based on the user’s historical behavior data and the corresponding product features and other information, requiring the user to have a clear target product.
  • the embodiments of the present disclosure are based on the user’s historical information, for example, including historical information.
  • the user’s shopping intent is filtered from the user’s historical subscription information, which is the pan-conceptual product information, and the pan-conceptual product generally refers to a certain There are three categories of goods, and the categories can be large or small.
  • a bag of rice For example, "a bag of rice”, “a bag of 10KG rice”, and “a bag of 10KGXX rice” are three categories from large to small, not a product with very clear demand. Therefore, when recommending products based on user needs, first obtain the user's historical information, and filter out the user's shopping intention from the historical information.
  • Step 102 Determine the key needs of the user according to the historical information
  • the user's shopping intention is filtered out of the above-mentioned user's historical information, which can be based on one piece of historical information or multiple product information, such as the user's recent historical browsing information, historical subscription information, and historical shopping
  • the search information is based on the type of household daily necessities, so the user’s shopping intention can also be called the user’s key needs.
  • the user’s demand range is determined by the user’s historical product subscription information, such as the rice information in daily necessities as the user’s history
  • the highest click-through rate, or the high frequency of historical searches, is determined to be a key user demand.
  • the user’s historical information is time-sensitive, that is, the user has historically browsed product information within a week or three days, or historically subscribed product information can be considered as real-time data that characterizes user needs, and the product information extracted from it is more user-friendly.
  • Products that are easy to buy, and the product information extracted from the historical subscription information is usually a large classification of the product, such as rice information in daily necessities, weight loss product information in health products, etc., which can be further based on the user’s historical browsing time for each specific product.
  • Historical information such as historical click rate or historical search rate further determines the key needs of users.
  • a weight value can be set for each historical information classification according to the user's preset importance of historical information in advance, and the key needs of the user can be calculated more efficiently and accurately based on the weight value and historical information.
  • Step 103 Extract candidate products matching the key requirements from a preset merchant list database
  • the user's demand range can be determined, such as the rice information in the daily necessities described above, or the weight-loss product information in the health products, and the user needs can be determined
  • the key demand is rice or weight-loss products.
  • Merchants who have purchased rice or weight-loss products can be further extracted from the preset merchant list database as candidate merchants, and the products sold by the candidate merchants are determined as conforming Candidates for critical needs.
  • the preset merchant list database usually stores the categories corresponding to all product information and the corresponding relationships with the merchants selling the products.
  • the embodiment of the present disclosure does not limit the form of storing data in the preset merchant list database.
  • Step 104 Extract commodity sales information of the candidate commodity
  • the sales information of the products for the determined key needs is obtained, for example, the specific sales information of rice or weight-loss products at each merchant, including the selling price, the unit of measurement, and the current discount rate. And historical sales records.
  • Step 105 According to the commodity sales information, obtain a commodity sales cost performance index that meets the key demand;
  • the cost performance of each product can be obtained, that is, a cost performance index is determined according to the selling price, popularity, discount rate, etc., so as to clearly and accurately express the value of each product.
  • the actual selling price is the price-performance index.
  • the cost performance index can also be determined according to the user’s preferences. For example, the user wants the fastest delivery speed or the highest historical evaluation. Therefore, when determining the product cost performance index, it is not limited to the above-described factors such as selling price, popularity, discount rate, etc. There is no limitation on this embodiment of the present disclosure.
  • Step 106 According to the commodity sales cost performance index, display the commodity for the commodity sales information to the user;
  • all candidate merchants are sorted according to the level of the above-determined cost-effectiveness index, where the higher cost-effectiveness index is ranked first, and the candidate merchant recommendation list is finally obtained.
  • the target product is displayed to the user in the form of a recommended list of candidate merchants, or only the merchants and products with the first preset digits are selected and displayed to the user.
  • the preset digits are selected by the user or set according to system requirements. Compare the embodiments of the present disclosure No restrictions.
  • the user's historical information is obtained; the key needs of the user are determined according to the historical information; the candidate products that match the key needs are extracted from the preset merchant list database; the candidate products are extracted According to the commodity sales information, obtain a commodity sales cost-effective index that meets the key requirement; according to the commodity sales cost-effective index, show the user the commodity for the commodity sales information. It has the ability to effectively understand the general needs in the user's historical information, provide users with the most cost-effective products according to the general needs of the users, and reduce the time-consuming beneficial effects of further screening and comparison of the users in the same category of products.
  • FIG. 2 a flowchart of steps of an embodiment of a recommendation method of the present disclosure is shown, which may specifically include the following steps:
  • Step 201 Obtain historical information of the user
  • This step is the same as step 101 and will not be described in detail here.
  • Step 202 The historical information includes one or more of historical browsing information, historical search information, historical subscription information, historical consumption information, and historical consumption cycles. Then, according to the historical information, determine the key needs of the user The steps include:
  • Sub-step S1 obtaining browsing keywords and historical browsing volume corresponding to the browsing keywords according to the historical browsing information
  • Sub-step S2 obtaining search keywords and historical search frequencies corresponding to the search keywords according to the historical search information
  • Sub-step S3 obtaining subscription keywords and historical reading frequencies corresponding to the subscription keywords according to the historical search information
  • Sub-step S4 acquiring consumption keywords and historical consumption corresponding to the consumption keywords according to the historical consumption information
  • each item of historical information mentioned above can be used to obtain respective information and user information.
  • Keywords related to consumer demand such as browse keywords, search keywords, subscription keywords, and consumer keywords.
  • the extraction of keywords can be extracted from massive amounts of historical data by using a pre-trained machine learning model.
  • the method for extracting keywords of respective historical information is not limited in the embodiment of the present disclosure.
  • Sub-step S5 comparing the consumption product corresponding to the historical consumption period with one or more of the browsing keyword, the search keyword, the subscription keyword, and the consumption keyword to obtain all Key words describing the consumer's demand
  • the historical consumption cycle is the rule that users purchase goods periodically. For example, if a user buys brand A washing powder every three months, then three months is the consumption cycle of brand A washing powder for the user.
  • the corresponding consumption keywords obtained in the user's historical consumption cycle are washing powder and brand A. Combine the above keywords with other historical keywords, such as browsing keywords, search keywords, subscription keywords, and one or more of consumption keywords By comparison, the key demand for user consumption is that they like high-cost, domestic brands and discounted brand A detergents.
  • Sub-step S6 according to the historical browsing volume, the historical search frequency, the historical reading frequency, and the historical consumption volume, the browsing keywords, the search keywords, the subscription keywords, and the The key weight value of consumer keywords;
  • the key weight values of historical keywords extracted from each historical information are respectively set according to the user's advancement of historical browsing volume, historical search frequency, historical reading frequency, and historical consumption.
  • the user subscribes to the keyword to set a higher weight value. If the consumer keyword is obtained for three historical information keywords, then the sum of the keyword weights of the three keywords is 1. The weight value of each keyword is set as the ratio of the usage rate of the respective historical information to the time of the user's overall consumption-related behavior.
  • the setting of the keyword weight value is not limited to the above description, and the embodiment of the present disclosure does not limit this.
  • Sub-step S7 sort the consumption demand keywords according to the size of the key weight value
  • the consumer keywords are sorted according to the weight values. For example, a number of key needs are extracted from the user’s historical information, such as washing powder under the category of daily necessities, milk powder in maternal and child products, and TV cabinets in furniture products.
  • the historical click rate of washing powder is the highest
  • milk powder is the highest
  • the historical subscription frequency of the TV cabinet is the highest.
  • the weight values of the three consumer keywords of washing powder, milk powder and TV cabinet are obtained, and the overall weight of the milk powder is the highest. Milk powder ranked first, and so on.
  • sub-step S8 the consumption keywords with the first preset digits in the ranking are selected and determined as the key needs of the user.
  • the top N positions are determined as the key needs of the user.
  • only the first one may be determined as the consumer keywords of the user. This embodiment of the present disclosure does not Be restricted.
  • Step 203 Extract candidate products matching the key requirements from a preset merchant list database
  • step 203 further includes:
  • Sub-step 2031 matching one or more of the browsing keywords, the search keywords, the subscription keywords, and the consumption keywords with a product catalog in a preset product information database to obtain a match Candidates for the key demand.
  • the keywords extracted from the user's historical product subscription information can be product feature words and product attribute information.
  • the product feature words are product name information
  • the product attribute information is product classification, brand name, and product measurement unit, etc.
  • the candidate product determined according to the above keywords is "a bag of 10KG Arowana Rice”.
  • Step 204 Extract commodity sales information of the candidate commodity
  • the candidate products extracted from the historical information are matched with the product attribute tags in the preset product information database.
  • the keyword extracted from the user's historical information is "large-capacity oil-control shampoo”.
  • These labels such as “volume”, “oil control”, etc.
  • the commodity attribute labels in the preset commodity information database have common effects of shampoo such as “oil control”, “dandruff”, “relief itching”, etc.
  • the net content range is [80ml ,250ml), [250ml,500ml), [500ml,750ml), [750ml,1L), etc.
  • the matching result is that the effect is "oil control” and the shampoo with the capacity "[750ml,1L)" is a candidate product that meets the key requirements , And then further obtain its sales information.
  • candidate merchant 1 sells the product at a price of 98, the unit of measurement is 800ml, and the current discount rate is 5% off.
  • the historical sales record is 1,000 monthly sales; the selling price of the product sold by candidate merchant 2 is 95, and the unit of measurement is 800ml.
  • the current discount rate is a gift of 150ml small package of similar shampoo.
  • the historical sales record is monthly sales of 500.
  • the unit price is different.
  • Step 205 Obtain a preset weight value selected by the user, where the preset weight value includes a cost performance index, a normalized unit measurement discount price, and a weight ratio among discount rate factors.
  • users can choose preset weight values according to specific needs, including the different importance of cost-effectiveness, product popularity, and discount rates to users, that is, users based on the differences in cost-effectiveness, product popularity, and discount rates. Choose the ratio between them for importance, and the sum of cost performance, product popularity, and discount rate is 1.
  • the proportions of cost performance, product popularity, and discount rate are 80%, 10%, and 10%, respectively, and their weight values are 0.8, 0.1, and 0.1, respectively.
  • Step 206 Determine the normalized unit price of the commodity according to the ratio of the selling price of the commodity to the measurement unit of the commodity; the commodity sales information includes the commodity selling price, the commodity measurement unit, the current discount rate of the commodity, and the commodity history The sales record is based on the commodity sales information.
  • a product in a category that users subscribe to with high popularity, high discounts, and high cost performance is a user's favorite product. Therefore, in order to obtain the cost performance index of the product, you first need to obtain the unit price of the product, that is, according to the price of the product and the product. Calculate the unit price of the commodity, that is, the normalized unit price. For example, if a 10KG bag of Arowana rice is 75 yuan, then the normalized unit price of Arowana rice is 7.5 yuan/kg.
  • Step 207 Determine the normalized unit measurement discount price of the commodity according to the normalized unit price of measurement and the commodity discount rate;
  • the actual unit price of the target product can be calculated, that is, the normalized unit measurement discount price.
  • the normalized unit price of Jinlongyu brand rice is 7.5 yuan/kg
  • the discount rate of the store is 9.7 percent
  • the normalized unit measurement discount price of Jinlongyu brand rice is 7.275 yuan/kg.
  • Step 208 Determine the historical maximum discount rate of the product according to the historical sales record of the product
  • Step 209 Determine the commodity discount rate factor according to the ratio between the current discount rate of the commodity and the maximum discount rate in the history of the commodity;
  • the historical lowest discount rate factor D i of the product can be obtained:
  • Step 210 Determine a commodity sales price index corresponding to the key demand according to the sum of the commodity discount rate factor and the normalized unit measurement discount price.
  • R i P i /c i +H i +R i
  • P i is the actual selling price of the commodity i
  • c i is the unit measurement that can be normalized for the commodity i.
  • P i /c i is the unit price of the commodity in the unit measurement, that is, the normalized unit of measurement
  • the price such as "Qingfeng log pure family paper 130 draws/pack 3.5 yuan"
  • P is 3.5 yuan
  • C is 130 draws, 0.0269 yuan/draw
  • H i is the sales volume of i during the period t, so calculated
  • the obtained R i is the cost performance index of the candidate merchant.
  • step 210 further includes:
  • Sub-step 2101 Calculate the weighted sum of the commodity discount rate factor and the normalized unit measurement discount price according to the weight ratio between the cost performance index, the normalized unit measurement discount unit price, and the discount rate factor, and determine it as The commodity sales cost index corresponding to the key demand.
  • the cost-effective index R i of the candidate merchant can be Calculate according to the following formula:
  • R i ⁇ P i /c i + ⁇ H i + ⁇ R i
  • the weighted sum of the discount rate factor of the target commodity and the discounted price measured by the normalized unit is calculated, and the obtained cost performance index takes into account the different importance of cost performance, commodity popularity, and discount rate to users.
  • the weight ratio of cost performance, product popularity, and discount rate is not limited to the way described in step 208 to be determined by the user's active selection, and can also be obtained by the system based on the user's historical shopping records. The next time the user makes a purchase, it is directly determined by the user The system makes a user-friendly choice, which is not limited by the embodiment of the present disclosure.
  • Step 211 Sort the commodity sales information according to the cost performance index to obtain a ranking list of candidate commodities
  • all candidate products are sorted according to the level of the above-determined cost performance index, where the higher cost index is ranked first, and the candidate product ranking list is finally obtained.
  • Step 212 Show the user a preset number of commodities in the candidate commodity ranking list.
  • the candidate product recommendation list is displayed to the user, or only the merchants and products with the first preset digits are selected to be displayed to the user.
  • the preset digits are selected by the user or set according to the system requirements. Compared with the implementation of the present invention The examples are not limited.
  • the historical information of the user is obtained; the historical information includes one or more of historical browsing information, historical search information, historical subscription information, historical consumption information, and historical consumption cycle; according to the historical Information to determine the key needs of the user; extract candidate products matching the key needs from a preset merchant list database; extract product sales information of the candidate products; obtain according to the product sales information that meets the key needs
  • the commodity sales price index; the commodity sales information includes commodity price, commodity measurement unit, commodity current discount rate, and commodity historical sales records; according to the commodity sales price index, the user is shown the commodity sales information commodity.
  • FIG. 3 there is shown a structural block diagram of an embodiment of a recommending device of the present disclosure, which may specifically include the following modules:
  • the historical information acquisition module 301 is used to acquire historical information of the user
  • the key requirement determining module 302 is configured to determine the key requirement of the user according to the historical information
  • the candidate product extraction module 303 is configured to extract candidate products that match the key demand from a preset merchant list database
  • the commodity sales information extraction module 304 is configured to extract commodity sales information of the candidate commodity
  • the cost performance index determining module 305 is configured to obtain, according to the commodity sales information, a commodity sales price index that meets the key demand;
  • the commodity display module 306 is configured to display the commodity for the commodity sales information to the user according to the commodity sales cost performance index.
  • the user's historical information is obtained; the key needs of the user are determined according to the historical information; the candidate products that match the key needs are extracted from the preset merchant list database; the candidate products are extracted According to the commodity sales information, obtain a commodity sales cost-effective index that meets the key requirement; according to the commodity sales cost-effective index, show the user the commodity for the commodity sales information. It has the ability to effectively understand the general needs in the user's historical information, provide users with the most cost-effective products according to the general needs of users, and reduce the time-consuming beneficial effect of further screening and comparison of similar products by users.
  • FIG. 4 a structural block diagram of an embodiment of a recommending device of the present disclosure is shown, which may specifically include the following modules:
  • the historical information acquisition module 301 is used to acquire historical information of the user
  • the key requirement determining module 302 is configured to determine the key requirement of the user according to the historical information
  • the historical information includes one or more of historical browsing information, historical search information, historical subscription information, historical consumption information, and historical consumption cycle.
  • the key demand determination module 302 further includes:
  • the historical browsing information determining sub-module is used to obtain browsing keywords and the historical browsing volume corresponding to the browsing keywords according to the historical browsing information;
  • the historical search frequency obtaining sub-module is used to obtain search keywords and historical search frequencies corresponding to the search keywords according to the historical search information;
  • the historical reading frequency obtaining sub-module is used to obtain subscription keywords and historical reading frequencies corresponding to the subscription keywords according to the historical search information;
  • the historical consumption obtaining sub-module is used to obtain consumption keywords and historical consumption corresponding to the consumption keywords according to the historical consumption information;
  • the consumer demand keyword acquisition sub-module is used to obtain one or more of the browsing keywords, the search keywords, the subscription keywords, and the consumer keywords according to the consumer product corresponding to the historical consumption cycle Items are compared to obtain the consumer demand keywords of the user;
  • the key weight value determining sub-module is used to determine the browsing keywords, the search keywords, and the subscriptions respectively according to the historical browsing volume, the historical search frequency, the historical reading frequency, and the historical consumption Keywords, the key weight value of the consumer keywords;
  • the keyword ranking sub-module is used to rank the consumer demand keywords according to the size of the key weight value
  • the key requirement determination sub-module is used to select the consumption keywords of the first preset digits in the ranking and determine it as the key requirement of the user.
  • the candidate product extraction module 303 is configured to extract candidate products that match the key demand from a preset merchant list database
  • the candidate commodity extraction module 303 includes:
  • the candidate product extraction sub-module 3031 is used to combine one or more of the browsing keywords, the search keywords, the subscription keywords, and the consumption keywords with the product catalog in the preset product information database Perform matching to obtain candidate commodities that match the key needs.
  • the commodity sales information extraction module 304 is configured to extract commodity sales information of the candidate commodity
  • it also includes:
  • the weight ratio obtaining module is used to obtain a preset weight value selected by the user, the preset weight value including a cost performance index, a normalized unit measurement discount price, and a weight ratio among discount rate factors.
  • the cost performance index determining module 305 is configured to obtain, according to the commodity sales information, a commodity sales price index that meets the key demand;
  • the commodity sales information includes commodity selling price, commodity measurement unit, commodity current discount rate and commodity historical sales records
  • said cost performance index determining module 305 includes:
  • the normalized unit price determination sub-module 3051 is configured to determine the normalized unit price of the commodity according to the ratio of the selling price of the commodity to the unit of measurement of the commodity;
  • a normalized unit measurement discount price determination sub-module 3052 configured to determine the normalized unit measurement discount price of the commodity according to the normalized unit price and the commodity discount rate;
  • the historical maximum discount rate determination sub-module 3053 is configured to determine the historical maximum discount rate of the product according to the historical sales record of the product;
  • the discount rate factor determination submodule 3054 is configured to determine the commodity discount rate factor according to the ratio between the current discount rate of the commodity and the maximum discount rate in the history of the commodity;
  • the price/performance index determination submodule 3055 is configured to determine the commodity sales price/performance index corresponding to the key demand according to the sum of the commodity discount rate factor and the normalized unit measurement discount price.
  • the cost performance index determining submodule 3055 further includes:
  • the cost performance index determining unit is configured to calculate the weighted sum of the commodity discount rate factor and the normalized unit measurement discount price according to the weight ratio between the cost performance index, the normalized unit measurement discount unit price, and the discount rate factor , Determined as the commodity sales cost index corresponding to the key demand.
  • the commodity display module 306 is configured to display the commodity for the commodity sales information to the user according to the commodity sales cost performance index.
  • the commodity display module 306 further includes:
  • the candidate product ranking list obtaining submodule 3061 is configured to rank the product sales information according to the cost performance index to obtain a candidate product ranking list
  • the commodity display sub-module 3062 is configured to show the user a preset number of commodities in the candidate commodity ranking list.
  • the description is relatively simple, and for related parts, please refer to the part of the description of the method embodiment.
  • the embodiment of the present disclosure further provides an electronic device, including: a processor, a memory, and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program when the program is executed.
  • an electronic device including: a processor, a memory, and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program when the program is executed.
  • the embodiment of the present disclosure also provides a readable storage medium, such as a non-volatile readable storage medium.
  • a readable storage medium such as a non-volatile readable storage medium.
  • the historical information of the user is obtained; the historical information includes one or more of historical browsing information, historical search information, historical subscription information, historical consumption information, and historical consumption cycle
  • determine the key needs of the user determine the key needs of the user; extract candidate products matching the key needs in the preset merchant list database; extract the product sales information of the candidate products; obtain according to the product sales information
  • the commodity sales price index that meets the key requirements; the commodity sales information includes commodity price, commodity measurement unit, current discount rate of the commodity, and commodity historical sales records; according to the commodity sales price index, the user is shown to the user The commodity that describes the commodity sales information.
  • the description is relatively simple, and for related parts, please refer to the part of the description of the method embodiment.
  • the embodiments of the embodiments of the present disclosure may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present disclosure may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present disclosure may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.
  • computer-usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • FIG. 5 shows a computing processing device that can implement the recommendation method according to the present invention, such as the aforementioned electronic device.
  • the computing processing device traditionally includes a processor 1010 and a computer program product in the form of a memory 1020 or a machine-readable medium.
  • the memory 1020 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), EPROM, hard disk, or ROM.
  • the memory 1020 has a storage space 1030 for executing program codes 1031 (instructions) of any method steps in the above methods.
  • the storage space 1030 for program codes may include various program codes 1031 for implementing various steps in the above method. These program codes can be read out from or written into one or more computer program products.
  • These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards or floppy disks.
  • Such a computer program product is usually a portable or fixed storage unit as described with reference to FIG. 6.
  • the storage unit may have storage segments, storage spaces, etc. arranged similarly to the memory 1020 in the computing processing device of FIG. 3.
  • the program code can be compressed in an appropriate form, for example.
  • the storage unit includes computer-readable codes 1031', that is, codes that can be read by, for example, a processor such as 1010. These codes, when run by a computing processing device, cause the computing processing device to execute the method described above. The various steps.
  • These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing terminal equipment to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction device.
  • the instruction device implements the functions specified in one process or multiple processes in the flowchart and/or one block or multiple blocks in the block diagram.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Development Economics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种推荐方法、装置、电子设备及可读存储介质,包括:获取用户的历史信息(101);根据所述历史信息,确定所述用户的关键需求(102);在预设商家列表数据库中提取匹配所述关键需求的候选商品(103);提取所述候选商品的商品销售信息(104);根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数(105);根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品(106)。解决了为现有技术根据用户需求推荐的商品还需要用户进一步筛选的问题。

Description

一种推荐方法、电子设备及可读存储介质
相关申请的交叉引用
本专利申请要求于2019年4月23日提交的、申请号为201910330329.7、发明名称为“一种推荐方法、装置、电子设备及可读存储介质”的中国专利申请的优先权,该申请的全文以引用的方式并入本文中。
技术领域
本公开涉及推荐技术领域,特别是涉及一种推荐方法、一种推荐装置、电子设备及可读存储介质
背景技术
根据用户提供的搜索行为信息获取用户的消费倾向,并为用户推荐对应的产品是目前广泛应用的商品推荐方式。
现有技术中,通过用户关注商品进行信息推荐的技术方案中,一般基于用户的历史行为数据、商品热度等信息进行推荐,缺点一方面是商品泛化性能比较差,只能推荐用户浏览过或者热度比较高的商品,而目前用户往往给出一个泛概念的商品(所谓泛概念商品一般是指某种类别商品,类别可大可小,例如“一袋大米”、“一袋10KG大米”、“一袋10KGXX牌大米”就是从大到小的三个类别),而不是一个需求十分明确的商品;另一方面,用户比较热衷于那些人气高、折扣率大且性价比高的商品,但目前商家促销手段多种多样,用户很难在不同商家选择出一个真正实惠且人气高的商品。
发明内容
鉴于上述问题,提出了本公开实施例以便提供一种克服上述问题或者至少部分地解决上述问题的一种推荐方法和相应的一种推荐装置。
根据本公开的第一方面,本公开实施例公开了一种推荐方法,具体包括:获取用户的历史信息;根据所述历史信息,确定所述用户的关键需求;在预设商家列表数据库中提取匹配所述关键需求的候选商品;提取所述候选商品的商品销售信息;根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。
根据本公开的第二方面,本公开实施例公开了一种推荐装置,具体包括:历史信息获取模块,用于获取用户的历史信息;关键需求确定模块,用于根据所述历史信息,确定所述用户的关键需求;候选商品提取模块,用于在预设商家列表数据库中提取匹配所述关键需求的候选商品;商品销售信息提取模块,用于提取所述候选商品的商品销售信息;性价比指数确定模块,用于根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;商品展示模块,用于根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。
根据本公开的第三方面,提供了一种电子设备,包括:处理器、存储器以及存储在所述存储器上并可在所述处理器上运行的计算机程序,其特征在于,所述处理器执行所述程序时实现如前述的推荐方法。
根据本公开的第四方面,提供了一种可读存储介质,当所述存储介质中的指令由电子设备的处理器执行时,使得电子设备能够实现前述的推荐方法。
本公开实施例包括以下优点:通过获取用户的历史信息,进而确定用户的泛需求,也即关键需求,找到符合关键需求对应商品的销售商家,将具备性价比较高的商品推荐给用户。具备有效理解用户的历史信息中的泛需求,根据用户的泛需求为用户提供性价比最高商品,减少用户在同类商品中进一步筛选比较导致耗费时间的有益效果。
附图说明
为了更清楚地说明本公开实施例的技术方案,下面将对本公开实施例的描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本公开的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。
图1是本公开实施例的一种推荐方法实施例的步骤流程图;
图2是本公开实施例的一种推荐方法实施例的步骤流程图;
图3是本公开实施例的一种推荐装置实施例的结构框图;
图4是本公开实施例的一种推荐装置实施例的结构框图;
图5示意性地示出了用于执行根据本公开实施例的方法的计算处理设备的框图;
图6示意性地示出了用于保持或者携带实现根据本公开实施例的方法的程序代码的存储单元。
具体实施方式
下面将结合本公开实施例中的附图,对本公开实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本公开一部分实施例,而不是全部的实施例。基于本公开中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本公开保护的范围。
实施例一
参照图1,示出了本公开的一种推荐方法实施例的步骤流程图,具体可以包括如下步骤:
步骤101,获取用户的历史信息;
在用户商品推荐领域中,通常基于用户的历史行为数据,以及对应的商品特度等信息进行推荐,要求用户有一个明确的目标商品,而本公开实施例是基于用户的历史信息,例如包含历史浏览信息、历史消费信息、历史订阅信息的一种或多种,以历史订阅信息为例,从用户历史订阅信息中筛选用户的购物意向,即为泛概念商品信息,而泛概念商品一般指某种类别商品,类别可大可小,例如“一袋大米”、“一袋10KG大米”、“一袋10KGXX牌大米”就是从大到小的三个类别,而不是一个需求十分明确的商品,所以针对用户需求推荐商品时,首先获取用户的历史信息,从历史信息中筛选出用户的购物意向。
步骤102,根据所述历史信息,确定所述用户的关键需求;
本公开实施例中,在上述用户的历史信息中筛选出用户的购物意向,可以是基于一个历史信息,也可以是基于多个商品信息,例如用户最近的历史浏览信息、历史订阅信息以及历史购物搜索信息都是基于家庭日用品的种类,那么可以将用户的购物意向,也可以称之为用户的关键需求,例如通过用户历史商品订阅信息确定用户的需求范围,如日用品中的大米信息为用户历史点击率最高,或历史搜索频率多高等,确定为用户的关键需求。
可以理解地,用户的历史信息具备时效性,即用户在一周内或者三天内历史浏览商品信息,或者历史订阅商品信息可以被认为是表征用户需求的实时数据,从中提取出的商品信息为用户较易购买的商品,而历史订阅信息中提取的商品信息通常为商品的大分类,例如日用品中的大米信息,保健品中的减肥产品信息等,可以进一步根据用户对各个具体产品的历史浏览时间,历史点击率或历史搜索率等历史信息,进一步确定用户的关键需求。
具体地,可以根据用户对历史信息的提前预设重要度,对每一个历史信息分类设置权重值,根据权重值与历史信息可以更高效和准确的计算出用户的关键需求。
步骤103,在预设商家列表数据库中提取匹配所述关键需求的候选商品;
本公开实施例中,例如通过历史订阅信息中提取商品的大分类信息后,可以确定用户的需求范围,如上述描述的日用品中的大米信息,或者保健品中的减肥产品信息,可以确定用户需求的关键需求为大米类或者减肥产品类,可以进一步在预设商家列表数据库中提取用户曾经购买过大米类或减肥产品类相关产品的商家,作为候选商家,并将候选商家销售的产品确定为符合关键需求的候选商品。
其中,预设商家列表数据库中通常会存储所有商品信息对应的分类,与销售该商品的商家的对应关系。对于预设商家列表数据库中存储数据的形式本公开实施例不加以限制。
步骤104,提取所述候选商品的商品销售信息;
本公开实施例中,在确定的候选商家中,获取针对确定的关键需求的商品的销售信息,例如,大米或者减肥产品在各商家的具体销售信息,其中包括售价、计量单位、当前折扣率以及历史销售记录。
当然,销售信息不限于上述描述,本公开实施例不加以限制。
步骤105,根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;
本公开实施例中,通过上述获取的各商家对于目标商品的销售信息,可以得出各商品的性价比,即根据售价、人气、折扣率等确定一个性价比指数,以便清晰准确的表达各商品的实际销售价格,即性价比指数。
其中,性价比指数还可以根据用户的喜好来确定,例如用户希望发货速度最快,或者历史评价最高等,所以在确定商品性价比指数时,不限于上述描述的售价、人气、折扣率等因素,对此本公开实施例不加以限制。
步骤106,根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品;
本公开实施例中,根据上述确定的性价比指数的高低,对所有候选商家进行排序,其中性价比指数高的排名靠前,最终获得候选商家推荐列表。将目标商品以候选商家推荐列表的形式展示给用户,或者只选取其中前预设位数的商家及商品展示给用户,预设位数由用户选择或者根据系统需求设定,对比本公开实施例不加以限制。
在本公开实施例中,通过获取用户的历史信息;根据所述历史信息,确定所述用户的关键需求;在预设商家列表数据库中提取匹配所述关键需求的候选商品;提取所述候选商品的商品销售信息;根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。具备有效理解用户的历史信息中的泛需求,根据用户的泛需求为用户提供性价比最高商品,减少用户在同 类商品中进一步筛选比较导致耗费时间的有益效果。
实施例二
参照图2,示出了本公开的一种推荐方法实施例的步骤流程图,具体可以包括如下步骤:
步骤201,获取用户的历史信息;
此步骤与步骤101相同,在此不再详述。
步骤202,所述历史信息包括历史浏览信息、历史搜索信息、历史订阅信息、历史消费信息、历史消费周期其中的一项或多项,则根据所述历史信息,确定所述用户的关键需求的步骤,包括:
子步骤S1,根据所述历史浏览信息获取浏览关键词以及对应所述浏览关键词的历史浏览量;
和/或,
子步骤S2,根据所述历史搜索信息获取搜索关键词以及对应所述搜索关键词的历史搜索频率;
和/或,
子步骤S3,根据所述历史搜索信息获取订阅关键词以及对应所述订阅关键词的历史阅读频率;
和/或,
子步骤S4,根据所述历史消费信息获取消费关键词以及对应所述消费关键词的历史消费量;
具体地,当历史信息包括历史浏览信息、历史搜索信息、历史订阅信息、历史消费信息、历史消费周期其中的一项或多项时,通过上述每一项历史信息,可以分别获取各自的与用户消费需求有关的关键词,如浏览关键词、搜索关键词、订阅关键词、消费关键词。
当然,关键词的提取可以利用预先训练好的机器学习模型从海量的历史数据中提取出来,对于各自历史信息关键词的提取方法,本公开实施例对此不加以限制。
子步骤S5,根据所述历史消费周期对应的消费产品与所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词中的一项或多项进行对比,得到所述用户的消费需求关键词;
具体地,历史消费周期为用户周期性购买商品的规律,例如,用户每隔三个月会购买A品牌的洗衣粉,那么三个月为A品牌洗衣粉针对该用户的消费周期,如此,在用户历史消费周期中获取的对应消费关键词为洗衣粉和A品牌,将上述关键词与其他历史关键词,如浏览 关键词、搜索关键词、订阅关键词、消费关键词的一项或多项进行对比,得到用户消费的关键需求为,喜欢性价比高、国产品牌、打折的A品牌的洗衣粉。
子步骤S6,根据所述历史浏览量、所述历史搜索频率、所述历史阅读频率、所述历史消费量分别确定所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词的关键权重值;
具体地,根据用户提前对历史浏览量、历史搜索频率、历史阅读频率、历史消费量的高低,分别设置各历史信息中提取出的历史关键词的关键权重值。
例如,用户阅读频率较高,那么用户订阅关键词则设置较高的权重值,如果消费关键词是针对三个历史信息关键词获取的,那么三个关键词的关键词权重的和为1,各自关键词的权重值分别设置为各自历史信息的使用率所占用户整体消费相关行为产生的时间的比值。
当然,关键词权重值的设置不限于上述描述,本公开实施例对此不加以限制。
子步骤S7,根据所述关键权重值的大小对所述消费需求关键词进行排序;
具体地,当得到各自消费需求关键词权重值后,根据权重值对各消费关键词进行排序。例如,从用户的历史消息中提取出多个关键需求,有日用品分类下的洗衣粉、有母婴用品中的奶粉、有家具用品中的电视柜等,而洗衣粉的历史点击率最高,奶粉的历史搜索率最高,电视柜的历史订阅频率最高,根据各自的历史信息使用率,得到洗衣粉、奶粉和电视柜这三个消费关键词各自的权重值,得出奶粉的整体权重最高,那么奶粉排名第一,以此类推。
子步骤S8,选择所述排序中的前预设位数的消费关键词,确定为所述用户的关键需求。
具体地,在得到的消费关键词的排序中,将排序靠前的N位确定为用户的关键需求,当然,也可以只确定第一位作为用户的消费关键词,对此本公开实施例不加以限制。
步骤203,在预设商家列表数据库中提取匹配所述关键需求的候选商品;
优选地,步骤203,进一步包括:
子步骤2031,将所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词中的一项或多项与预设商品信息数据库中的商品目录进行匹配,得到匹配所述关键需求的候选商品。
具体地,例如,在用户的历史商品订阅信息中提取的关键词可以为商品特征词和商品属性信息,通常商品特征词为商品名称信息,商品属性信息为商品分类、品牌名称以及商品计量单位等,将商品名称信息与本地预设商品信息数据库商品类别信息进行匹配,得到符合关键需求的候选商品。
例如,历史订阅信息中包含金龙鱼大米,其中的商品属性为乳玉皇妃特级东北大米 10KG”,那么根据上述关键词确定的候选商品即为“一袋10KG金龙鱼牌大米”。
步骤204,提取所述候选商品的商品销售信息;
具体地,将在历史信息中提取的候选商品与预设商品信息数据库中的商品属性标签进行匹配,例如在用户历史信息中提取的关键词为是“大容量控油洗发水”,需要理解“大容量”、“控油”等这些标签,预设商品信息数据库中的商品属性标签中有洗发水下常见的功效有“控油”、“去屑”、“止痒”等,净含量区间有[80ml,250ml)、[250ml,500ml)、[500ml,750ml)、[750ml,1L)等,匹配结果得到功效为“控油”容量为“[750ml,1L)”的洗发水为符合关键需求的候选商品,再进一步获取其销售信息。
首先,在预设商家列表数据库中提取售卖控油、容量为“[750ml,1L)”的洗发水的商家作为候选商家。
可以理解地,还可以据用户的定位信息,为用户自动筛选地理范围为3KM,5KM且含有处于优惠活动的该类别商品的商家确定为候选商家。
除此之外,还可以考察用户历史数据,将用户经常消费的商家,且含有该优惠活动的商品作为候选商家。
当然,在确定候选商家时,可以同时考虑上述因素的一种或多种,对此本公开实施例不加以限制。
其次,获取各候选商家针对控油、容量为“[750ml,1L)”的洗发水的销售信息,例如,候选商家1的销售该商品的售价为98,计量单位800ml,当前折扣率9.5折,历史销售记录为月销量1000;候选商家2的销售该商品的售价为95,计量单位800ml,当前折扣率为赠送150ml小包装同类洗发水,历史销售记录为月销量500,而两种商品得到的单价是不同的。
步骤205,获取用户选择的预设权重值,所述预设权重值包括性价比指数、归一化单位计量折扣价格、折扣率因子之间的权重比例。
具体地,在候选商家中,用户可以根据具体需求进行预设权重值的选择,其中包括性价比、商品人气、折扣率对在用户的不同重要度,即用户根据性价比、商品人气、折扣率的不同重要度选择它们之间的比值,且性价比、商品人气、折扣率总和为1。
例如,性价比、商品人气、折扣率的比重分别为80%,10%和10%,其权重值分别为0.8、0.1、0.1。
步骤206,根据所述商品售价与所述商品计量单位的比值,确定所述商品归一化计量单位价格;所述商品销售信息包括商品售价、商品计量单位、商品当前折扣率以及商品历史销售记录,根据所述商品销售信息。
具体地,通常人气高、折扣大、性价比高的用户订阅的类别下的某商品为用户的热衷商品,所以为了获取商品的性价比指数,首先需要获取商品在单价,即根据商品的售价与商品的计量单位的比值,计算出商品单价,即归一化计量单位价格,例如,一袋10KG金龙鱼牌大米为75元,那么金龙鱼牌大米归一化计量单位价格为7.5元/千克。
步骤207,根据所述归一化计量单位价格与所述商品折扣率,确定所述商品归一化单位计量折扣价格;
其中,根据当前目标商品的折扣率,可以计算出目标商品的实际单价,即归一化单位计量折扣价格。例如金龙鱼牌大米归一化计量单位价格为7.5元/千克,该店折扣率为9.7折,那么金龙鱼牌大米归一化单位计量折扣价格为7.275元/千克。
步骤208,根据所述商品历史销售记录,确定所述商品历史最大折扣率;
步骤209,根据所述商品当前折扣率与所述商品历史最大折扣率之间的比值,确定所述商品折扣率因子;
具体地,获取目标商品的历史销售记录中,可以得到该商品的历史最低折扣率因子D i
D i=MAX(d i_H max/d i,d o_max/d i)
其中d i为商品i的折扣率,d i_H max为t这段时间当前商品的最大折扣率,d o_max为备选泛概念商品列表的最大折扣率,如当前商品对外宣传“8.5折”促销,时间段t内商品i的最高折扣率为8折,而备选商品列表最大折扣率为7折,D i为8/8.5=0.941;
步骤210,根据所述商品折扣率因子与所述归一化单位计量折扣价格的和,确定对应所述关键需求的商品销售性价比指数。
具体地,定义商品的性价比指数为R i,可以根据如下公式进行计算:
R i=P i/c i+H i+R i
其中P i是该商品i的实际售价,c i是该商品i可通用归一化的单位计量,则P i/c i即为该单位计量上商品的单价,即,归一化计量单位价格,如“清风原木纯品家庭装抽纸130抽/包3.5元”P即为3.5元,C为130抽,为0.0269元/抽;H i为t这段时间i的销售量,如此计算得出的R i即为该候选商家的性价比指数。
优选地,步骤210,进一步包括:
子步骤2101,根据所述性价比指数、归一化单位计量折扣单价、折扣率因子之间的权重 比例,计算所述商品折扣率因子与所述归一化单位计量折扣价格的加权和,确定为对应所述关键需求的所述商品销售性价比指数。
具体地,如步骤208描述的,如果添加了性价比指数、归一化单位计量折扣单价、折扣率因子之间的权重比例,分别为α、β、γ,那么候选商家的性价比指数R i,可以根据如下公式进行计算:
R i=αP i/c i+βH i+γR i
其中,计算了目标商品的折扣率因子与所述归一化单位计量折扣价格的加权和,所得到的性价比指数考虑了性价比、商品人气、折扣率对用户的不同重要度。
可以理解地,性价比、商品人气、折扣率的权重比值不限于步骤208中描述的由用户主动选择的方式确定,也可以由系统根据用户的历史购物记录分析得到,在下次用户购物时,直接由系统做出针对于用户有利的选择,对此本公开实施例不加以限制。
步骤211,根据所述性价比指数,将所述商品销售信息进行排序,得到候选商品排序列表;
本公开实施例中,根据上述确定的性价比指数的高低,对所有候选商品进行排序,其中性价比指数高的排名靠前,最终获得候选商品排序列表。
步骤212,向所述用户展示所述候选商品排序列表中预设位数的商品。
本公开实施例中,将候选商品推荐列表展示给用户,或者只选取其中前预设位数的商家及商品展示给用户,预设位数由用户选择或者根据系统需求设定,对比本发明实施例不加以限制。
在本公开实施例中,通过获取用户的历史信息;所述历史信息包括历史浏览信息、历史搜索信息、历史订阅信息、历史消费信息、历史消费周期其中的一项或多项;根据所述历史信息,确定所述用户的关键需求;在预设商家列表数据库中提取匹配所述关键需求的候选商品;提取所述候选商品的商品销售信息;根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;所述商品销售信息包括商品售价、商品计量单位、商品当前折扣率以及商品历史销售记录;根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。具备有效理解用户的订阅内容中对泛需求概念的理解,为用户提供性价比最高商家,减少用户在同类商家中再次进行性价比比较的导致耗费时间的有益效果。
需要说明的是,对于方法实施例,为了简单描述,故将其都表述为一系列的动作组合,但是本领域技术人员应该知悉,本公开实施例并不受所描述的动作顺序的限制,因为依据本 公开实施例,某些步骤可以采用其他顺序或者同时进行。其次,本领域技术人员也应该知悉,说明书中所描述的实施例均属于优选实施例,所涉及的动作并不一定是本公开实施例所必须的。
实施例三
参照图3,示出了本公开的一种推荐装置实施例的结构框图,具体可以包括如下模块:
历史信息获取模块301,用于获取用户的历史信息;
关键需求确定模块302,用于根据所述历史信息,确定所述用户的关键需求;
候选商品提取模块303,用于在预设商家列表数据库中提取匹配所述关键需求的候选商品;
商品销售信息提取模块304,用于提取所述候选商品的商品销售信息;
性价比指数确定模块305,用于根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;
商品展示模块306,用于根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。
在本公开实施例中,通过获取用户的历史信息;根据所述历史信息,确定所述用户的关键需求;在预设商家列表数据库中提取匹配所述关键需求的候选商品;提取所述候选商品的商品销售信息;根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。具备有效理解用户的历史信息中的泛需求,根据用户的泛需求为用户提供性价比最高商品,减少用户在同类商品中进一步筛选比较导致耗费时间的有益效果。
实施例四
参照图4,示出了本公开的一种推荐装置实施例的结构框图,具体可以包括如下模块:
历史信息获取模块301,用于获取用户的历史信息;
关键需求确定模块302,用于根据所述历史信息,确定所述用户的关键需求;
优选地,所述历史信息包括历史浏览信息、历史搜索信息、历史订阅信息、历史消费信息、历史消费周期其中的一项或多项,所述关键需求确定模块302,进一步包括:
历史浏览信息确定子模块,用于根据所述历史浏览信息获取浏览关键词以及对应所述浏览关键词的历史浏览量;
和/或,
历史搜索频率获取子模块,用于根据所述历史搜索信息获取搜索关键词以及对应所述搜索关键词的历史搜索频率;
和/或,
历史阅读频率获取子模块,用于根据所述历史搜索信息获取订阅关键词以及对应所述订阅关键词的历史阅读频率;
和/或,
历史消费量获取子模块,用于根据所述历史消费信息获取消费关键词以及对应所述消费关键词的历史消费量;
消费需求关键词获取子模块,用于根据所述历史消费周期对应的消费产品与所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词中的一项或多项进行对比,得到所述用户的消费需求关键词;
关键权重值确定子模块,用于根据所述历史浏览量、所述历史搜索频率、所述历史阅读频率、所述历史消费量分别确定所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词的关键权重值;
关键词排序子模块,用于根据所述关键权重值的大小对所述消费需求关键词进行排序;
关键需求确定子模块,用于选择所述排序中的前预设位数的消费关键词,确定为所述用户的关键需求。
候选商品提取模块303,用于在预设商家列表数据库中提取匹配所述关键需求的候选商品;
优选地,所述候选商品提取模块303,包括:
候选商品提取子模块3031,用于将所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词中的一项或多项与预设商品信息数据库中的商品目录进行匹配,得到匹配所述关键需求的候选商品。
商品销售信息提取模块304,用于提取所述候选商品的商品销售信息;
优选地,还包括:
权重比例获取模块,用于获取用户选择的预设权重值,所述预设权重值包括性价比指数、归一化单位计量折扣价格、折扣率因子之间的权重比例。
性价比指数确定模块305,用于根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;
优选地,所述商品销售信息包括商品售价、商品计量单位、商品当前折扣率以及商品历史销售记录,所述性价比指数确定模块305,包括:
归一化计量单位价格确定子模块3051,用于根据所述商品售价与所述商品计量单位的比值,确定所述商品归一化计量单位价格;
归一化单位计量折扣价格确定子模块3052,用于根据所述归一化计量单位价格与所述商品折扣率,确定所述商品归一化单位计量折扣价格;
历史最大折扣率确定子模块3053,用于根据所述商品历史销售记录,确定所述商品历史最大折扣率;
折扣率因子确定子模块3054,用于根据所述商品当前折扣率与所述商品历史最大折扣率之间的比值,确定所述商品折扣率因子;
性价比指数确定子模块3055,用于根据所述商品折扣率因子与所述归一化单位计量折扣价格的和,确定对应所述关键需求的商品销售性价比指数。
优选地,所述性价比指数确定子模块3055,进一步包括:
性价比指数确定单元,用于根据所述性价比指数、归一化单位计量折扣单价、折扣率因子之间的权重比例,计算所述商品折扣率因子与所述归一化单位计量折扣价格的加权和,确定为对应所述关键需求的所述商品销售性价比指数。
商品展示模块306,用于根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。
优选地,所述商品展示模块306,进一步包括:
候选商品排序列表获取子模块3061,用于根据所述性价比指数,将所述商品销售信息进行排序,得到候选商品排序列表;
商品展示子模块3062,用于向所述用户展示所述候选商品排序列表中预设位数的商品。
对于装置实施例而言,由于其与方法实施例基本相似,所以描述的比较简单,相关之处参见方法实施例的部分说明即可。
本公开实施例还提供一种电子设备,包括:处理器、存储器以及存储在所述存储器上并可在所述处理器上运行的计算机程序,其特征在于,所述处理器执行所述程序时实现如上述的一个或多个所述的推荐方法。
本公开实施例还提供一种可读存储介质,例如非易失性可读存储介质。当所述非易失性可读存储介质中的指令由电子设备的处理器执行时,使得电子设备能够执行如所述的推荐方法。
综上所述,在本公开实施例中,通过获取用户的历史信息;所述历史信息包括历史浏览信息、历史搜索信息、历史订阅信息、历史消费信息、历史消费周期其中的一项或多项;根据所述历史信息,确定所述用户的关键需求;在预设商家列表数据库中提取匹配所述关键需求的候选商品;提取所述候选商品的商品销售信息;根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;所述商品销售信息包括商品售价、商品计量单位、商品当前折扣率以及商品历史销售记录;根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。具备有效理解用户的订阅内容中对泛需求概念的理解,为用户提供性价比最高商家,减少用户在同类商家中再次进行性价比比较的导致耗费时间的有益效果。
其具有如下优点:
一.在更少的交互次数基础上,只是通过用户的历史信息理解用户的购物关键需求,提高交互体验;
二.能够在有限的推送次数内,为用户发现令用户感兴趣的商品,提高购买量;
三.适用于线上线下不同情形,不需要过多的用户历史信息,适应程度广。
对于装置实施例而言,由于其与方法实施例基本相似,所以描述的比较简单,相关之处参见方法实施例的部分说明即可。
本说明书中的各个实施例均采用递进的方式描述,每个实施例重点说明的都是与其他实施例的不同之处,各个实施例之间相同相似的部分互相参见即可。
本领域内的技术人员应明白,本公开实施例的实施例可提供为方法、装置、或计算机程序产品。因此,本公开实施例可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且,本公开实施例可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
例如,图5示出了可以实现根据本发明的推荐方法的计算处理设备,例如前述的电子设备。该计算处理设备传统上包括处理器1010和以存储器1020形式的计算机程序产品或者机器可读介质。存储器1020可以是诸如闪存、EEPROM(电可擦除可编程只读存储器)、EPROM、硬盘或者ROM之类的电子存储器。存储器1020具有用于执行上述方法中的任何方法步骤的程序代码1031(指令)的存储空间1030。例如,用于程序代码的存储空间1030可以包括分别用于实现上面的方法中的各种步骤的各个程序代码1031。这些程序代码可以从一个或者多个计算机程序产品中读出或者写入到这一个或者多个计算机程序产品中。这些计算机程序产 品包括诸如硬盘,紧致盘(CD)、存储卡或者软盘之类的程序代码载体。这样的计算机程序产品通常为如参考图6所述的便携式或者固定存储单元。该存储单元可以具有与图3的计算处理设备中的存储器1020类似布置的存储段、存储空间等。程序代码可以例如以适当形式进行压缩。通常,存储单元包括计算机可读代码1031’,即可以由例如诸如1010之类的处理器读取的代码,这些代码当由计算处理设备运行时,导致该计算处理设备执行上面所描述的方法中的各个步骤。
本公开实施例是参照根据本公开实施例的方法、终端设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理终端设备的处理器以产生一个机器,使得通过计算机或其他可编程数据处理终端设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理终端设备以特定方式工作的计算机可读存储器中,使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品,该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理终端设备上,使得在计算机或其他可编程终端设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程终端设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
尽管已描述了本公开实施例的优选实施例,但本领域内的技术人员一旦得知了基本创造性概念,则可对这些实施例做出另外的变更和修改。所以,所附权利要求意欲解释为包括优选实施例以及落入本公开实施例范围的所有变更和修改。
最后,还需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者终端设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者终端设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不 排除在包括所述要素的过程、方法、物品或者终端设备中还存在另外的相同要素。
以上对本公开所提供的一种推荐方法、装置、电子设备及可读存储介质,进行了详细介绍,本文中应用了具体个例对本公开的原理及实施方式进行了阐述,以上实施例的说明只是用于帮助理解本公开的方法及其核心思想;同时,对于本领域的一般技术人员,依据本公开的思想,在具体实施方式及应用范围上均会有改变之处,综上所述,本说明书内容不应理解为对本公开的限制。

Claims (10)

  1. 一种推荐方法,其特征在于,包括:
    获取用户的历史信息;
    根据所述历史信息,确定所述用户的关键需求;
    在预设商家列表数据库中提取匹配所述关键需求的候选商品;
    提取所述候选商品的商品销售信息;
    根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;
    根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。
  2. 根据权利要求1所述的方法,其特征在于,所述历史信息包括历史浏览信息、历史搜索信息、历史订阅信息、历史消费信息、历史消费周期其中的一项或多项,所述根据所述历史信息,确定所述用户的关键需求的步骤,包括:
    根据所述历史浏览信息获取浏览关键词以及对应所述浏览关键词的历史浏览量;
    和/或,
    根据所述历史搜索信息获取搜索关键词以及对应所述搜索关键词的历史搜索频率;
    和/或,
    根据所述历史搜索信息获取订阅关键词以及对应所述订阅关键词的历史阅读频率;
    和/或,
    根据所述历史消费信息获取消费关键词以及对应所述消费关键词的历史消费量;
    根据所述历史消费周期对应的消费产品与所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词中的一项或多项进行对比,得到所述用户的消费需求关键词;
    根据所述历史浏览量、所述历史搜索频率、所述历史阅读频率、所述历史消费量分别确定所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词的关键权重值;
    根据所述关键权重值的大小对所述消费需求关键词进行排序;
    选择所述排序中的前预设位数的消费关键词,确定为所述用户的关键需求。
  3. 根据权利要求1所述的方法,其特征在于,所述商品销售信息包括商品售价、商品计量单位、商品当前折扣率以及商品历史销售记录,所述根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数的步骤,包括:
    根据所述商品售价与所述商品计量单位的比值,确定所述商品归一化计量单位价格;
    根据所述归一化计量单位价格与所述商品折扣率,确定所述商品归一化单位计量折扣价格;
    根据所述商品历史销售记录,确定所述商品历史最大折扣率;
    根据所述商品当前折扣率与所述商品历史最大折扣率之间的比值,确定所述商品折扣率因子;
    根据所述商品折扣率因子与所述归一化单位计量折扣价格的和,确定对应所述关键需求的商品销售性价比指数。
  4. 根据权利要求3所述的方法,其特征在于,所述根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品的步骤,包括:
    根据所述性价比指数,将所述商品销售信息进行排序,得到候选商品排序列表;
    向所述用户展示所述候选商品排序列表中预设位数的商品。
  5. 根据权利要求2所述的方法,其特征在于,所述在预设商家列表数据库中提取匹配所述关键需求的候选商品的步骤,包括:
    将所述浏览关键词、所述搜索关键词、所述订阅关键词、所述消费关键词中的一项或多项与预设商品信息数据库中的商品目录进行匹配,得到匹配所述关键需求的候选商品。
  6. 根据权利要求3所述的方法,其特征在于,在所述根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数的步骤之前,还包括:
    获取用户选择的预设权重值,所述预设权重值包括性价比指数、归一化单位计量折扣价格、折扣率因子之间的权重比例。
  7. 根据权利要求6所述的方法,其特征在于,所述根据所述商品折扣率因子与所述归一化单位计量折扣价格的和,确定对应所述关键需求的商品销售性价比指数的步骤,包括:
    根据所述性价比指数、归一化单位计量折扣单价、折扣率因子之间的权重比例,计算所述商品折扣率因子与所述归一化单位计量折扣价格的加权和,确定为对应所述关键需求的所述商品销售性价比指数。
  8. 一种推荐装置,其特征在于,包括:
    历史信息获取模块,用于获取用户的历史信息;
    关键需求确定模块,用于根据所述历史信息,确定所述用户的关键需求;
    候选商品提取模块,用于在预设商家列表数据库中提取匹配所述关键需求的候选商品;
    商品销售信息提取模块,用于提取所述候选商品的商品销售信息;
    性价比指数确定模块,用于根据所述商品销售信息,获取符合所述关键需求的商品销售性价比指数;
    商品展示模块,用于根据所述商品销售性价比指数,向所述用户展示针对所述商品销售信息的商品。
  9. 一种电子设备,其特征在于,包括:
    处理器、存储器以及存储在所述存储器上并可在所述处理器上运行的计算机程序,其特征在于,所述处理器执行所述程序时实现如权利要求1-7之任一项所述的推荐方法。
  10. 一种可读存储介质,其特征在于,当所述存储介质中的指令由电子设备的处理器执行时,使得电子设备能够实现如权利要求1-7之任一项所述的测推荐方法。
PCT/CN2019/125713 2019-04-23 2019-12-16 一种推荐方法、电子设备及可读存储介质 WO2020215751A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910330329.7 2019-04-23
CN201910330329.7A CN110197415A (zh) 2019-04-23 2019-04-23 一种推荐方法、装置、电子设备及可读存储介质

Publications (1)

Publication Number Publication Date
WO2020215751A1 true WO2020215751A1 (zh) 2020-10-29

Family

ID=67752003

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/125713 WO2020215751A1 (zh) 2019-04-23 2019-12-16 一种推荐方法、电子设备及可读存储介质

Country Status (2)

Country Link
CN (1) CN110197415A (zh)
WO (1) WO2020215751A1 (zh)

Families Citing this family (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110197415A (zh) * 2019-04-23 2019-09-03 北京三快在线科技有限公司 一种推荐方法、装置、电子设备及可读存储介质
CN110689402A (zh) * 2019-09-04 2020-01-14 北京三快在线科技有限公司 推荐商家的方法、装置、电子设备及可读存储介质
CN110533485A (zh) * 2019-09-05 2019-12-03 金瓜子科技发展(北京)有限公司 一种对象选取的方法、装置、存储介质及电子设备
CN110737694B (zh) * 2019-09-17 2022-09-23 中电万维信息技术有限责任公司 一种政务数据资源目录智能订阅方法
CN110992146A (zh) * 2019-12-02 2020-04-10 泰康保险集团股份有限公司 服务信息推送方法、装置、设备及存储介质
CN111080418A (zh) * 2019-12-28 2020-04-28 广东奥园奥买家电子商务有限公司 一种基于电商平台的商品组合销售方法和装置以及设备
CN111210308B (zh) * 2020-01-03 2024-03-26 北京秒针人工智能科技有限公司 一种推广策略的确定方法、装置、计算机设备和介质
CN111461782A (zh) * 2020-03-31 2020-07-28 广东奥园奥买家电子商务有限公司 一种电商商品的信息推送方法和装置以及设备
CN113762993A (zh) * 2020-06-03 2021-12-07 阿里巴巴集团控股有限公司 数据处理方法、系统、装置、设备和存储介质
CN111695971B (zh) * 2020-06-12 2024-01-12 腾讯科技(深圳)有限公司 物品推荐方法、装置和设备及计算机存储介质
CN112288464A (zh) * 2020-10-16 2021-01-29 杭州贝嘟科技有限公司 商品推荐方法、装置、计算机设备和存储介质
CN112419078A (zh) * 2020-12-03 2021-02-26 中国人寿保险股份有限公司 产品推荐方法、装置和电子设备
CN113159905A (zh) * 2021-05-20 2021-07-23 深圳马六甲网络科技有限公司 新用户的商品推荐方法、装置、设备及存储介质
CN113378057A (zh) * 2021-06-29 2021-09-10 珠海必要工业科技股份有限公司 一种信息提示方法、装置、计算机设备及存储介质
CN113297503A (zh) * 2021-07-26 2021-08-24 北京天健智慧科技有限公司 互联网护理信息平台的业务推荐方法
CN113919912B (zh) * 2021-10-13 2023-06-02 广州风雷益信息技术有限公司 一种产品智能选择方法、系统、设备及可读存储介质
CN114612126A (zh) * 2021-12-08 2022-06-10 江苏众亿国链大数据科技有限公司 一种基于大数据的信息推送方法
CN114240505B (zh) * 2021-12-21 2022-07-26 南京吐纳健康科技有限公司 一种基于大数据的空气净化器推荐系统及方法
CN115482065A (zh) * 2022-09-22 2022-12-16 保利和悦生活科技服务有限公司 基于大数据与电子商城平台的购物数据共享方法及系统
CN117035947B (zh) * 2023-10-09 2024-01-09 新立讯科技股份有限公司 基于大数据处理的农产品数据分析方法及云平台

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104599160A (zh) * 2015-02-06 2015-05-06 腾讯科技(深圳)有限公司 商品推荐方法和装置
CN106709741A (zh) * 2015-07-22 2017-05-24 中兴通讯股份有限公司 一种商品推荐方法及系统
CN106910108A (zh) * 2017-01-24 2017-06-30 武汉奇米网络科技有限公司 一种商品列表展示方法及系统
CN106933821A (zh) * 2015-12-29 2017-07-07 中国电信股份有限公司 一种基于相似度计算的个性化职位推荐方法和系统
CN107767227A (zh) * 2017-10-31 2018-03-06 深圳春沐源控股有限公司 网上商城的商品推荐方法、商品推荐系统
CN110197415A (zh) * 2019-04-23 2019-09-03 北京三快在线科技有限公司 一种推荐方法、装置、电子设备及可读存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104599160A (zh) * 2015-02-06 2015-05-06 腾讯科技(深圳)有限公司 商品推荐方法和装置
CN106709741A (zh) * 2015-07-22 2017-05-24 中兴通讯股份有限公司 一种商品推荐方法及系统
CN106933821A (zh) * 2015-12-29 2017-07-07 中国电信股份有限公司 一种基于相似度计算的个性化职位推荐方法和系统
CN106910108A (zh) * 2017-01-24 2017-06-30 武汉奇米网络科技有限公司 一种商品列表展示方法及系统
CN107767227A (zh) * 2017-10-31 2018-03-06 深圳春沐源控股有限公司 网上商城的商品推荐方法、商品推荐系统
CN110197415A (zh) * 2019-04-23 2019-09-03 北京三快在线科技有限公司 一种推荐方法、装置、电子设备及可读存储介质

Also Published As

Publication number Publication date
CN110197415A (zh) 2019-09-03

Similar Documents

Publication Publication Date Title
WO2020215751A1 (zh) 一种推荐方法、电子设备及可读存储介质
US11475501B2 (en) Recommendations based on branding
US10049139B2 (en) Diversity within search results
CN102609869B (zh) 商品购买系统和方法
CN112837118B (zh) 企业用户的商品推荐方法和装置
TWI512653B (zh) Information providing method and apparatus, method and apparatus for determining the degree of comprehensive relevance
CN103246980B (zh) 信息输出方法及服务器
US20040078214A1 (en) Product recommendation in a network-based commerce system
CN108121737A (zh) 一种业务对象属性标识的生成方法、装置和系统
JP2019512809A (ja) サービス対象のための時分割推薦方法及び装置
CN101887437A (zh) 一种搜索结果生成方法及信息搜索系统
WO2018090786A1 (zh) 一种网站站内精确搜索方法
WO2013016503A1 (en) Providing social product recommendations
JP2014500543A (ja) オンライン取引プラットフォームに基づくデータ処理
US20130254025A1 (en) Item ranking modeling for internet marketing display advertising
WO2017148272A1 (zh) 一种目标用户的识别方法和装置
CN110599307A (zh) 一种商品推荐的方法及装置
CN107633416A (zh) 一种业务对象的推荐方法、装置和系统
CN113689259A (zh) 基于用户行为的商品个性化推荐方法及系统
TWI552099B (zh) Recommended user method and device
JP2014115951A (ja) 属性情報最適化装置、属性情報最適化プログラム及び属性情報の最適化方法、並びにレコメンド対象選択装置、レコメンド対象選択プログラム及びレコメンド対象の選択方法
Prasetyo Searching cheapest product on three different e-commerce using k-means algorithm
CN110647679A (zh) 一种o2o场景下的搜索结果综合排序方法
CN112561581A (zh) 一种推荐方法、装置、电子设备及存储介质
Lin et al. A recommender system based on contextual information of click and purchase data to items for e-commerce

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19926686

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19926686

Country of ref document: EP

Kind code of ref document: A1