WO2022161116A1 - 物品展示方法和装置 - Google Patents

物品展示方法和装置 Download PDF

Info

Publication number
WO2022161116A1
WO2022161116A1 PCT/CN2022/070059 CN2022070059W WO2022161116A1 WO 2022161116 A1 WO2022161116 A1 WO 2022161116A1 CN 2022070059 W CN2022070059 W CN 2022070059W WO 2022161116 A1 WO2022161116 A1 WO 2022161116A1
Authority
WO
WIPO (PCT)
Prior art keywords
item
entity feature
feature information
entity
similarity
Prior art date
Application number
PCT/CN2022/070059
Other languages
English (en)
French (fr)
Inventor
杜继磊
Original Assignee
北京电解智科技有限公司
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 北京电解智科技有限公司 filed Critical 北京电解智科技有限公司
Publication of WO2022161116A1 publication Critical patent/WO2022161116A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION 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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0641—Shopping interfaces
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching

Definitions

  • the present disclosure relates to the field of computer technology, and in particular, to a method and apparatus for displaying items.
  • B2B Business-to-Business, business-to-business
  • B2C Business-to-Consumer, business-to-user
  • the embodiments of the present disclosure provide an article display method and device, which can improve the accuracy and display efficiency of displayed articles, improve the matching degree, reduce the calculation cost, and improve the user experience.
  • a method for displaying articles including:
  • the item information includes the item type and the item name;
  • the historical item set includes the entity feature information of multiple items ;
  • the similarity of the items is calculated, the item to be displayed is determined according to the similarity calculation result, and the item to be displayed is displayed.
  • the item information also includes an item picture
  • the method before the step of performing entity feature extraction on the item information corresponding to the first item, the method further includes:
  • Image recognition processing is performed on the image of the item to obtain the corresponding text information and item attribute information.
  • the method further includes a step of constructing the collection of historical items, including:
  • a collection of historical items is constructed based on entity feature information corresponding to multiple items.
  • the step of determining the second item from the historical item set according to the entity feature information corresponding to the first item includes:
  • the second item is determined from the historical item set according to the first weight coefficient of the entity feature information corresponding to the first item and the second item quantity threshold.
  • the step of calculating the similarity of the items includes:
  • Item similarity is calculated according to the first similarity of each entity feature information after entity alignment.
  • the step of calculating the similarity of the item also includes:
  • Item similarity is calculated according to the first similarity of each entity feature information after entity feature alignment and the second weight coefficient corresponding to each entity feature information.
  • an article display device comprising:
  • an information acquisition module configured to acquire item information corresponding to the first item, perform entity feature extraction on the item information corresponding to the first item, and obtain entity feature information corresponding to the first item; wherein the item information includes the item type and the item name;
  • the entity feature alignment module is used to determine the second item from the historical item set according to the entity feature information corresponding to the first item, and perform entity feature alignment on the entity feature information of the first item and the second item; wherein, the historical item set includes Entity feature information of multiple items;
  • the display module is used for calculating the similarity of the items according to the entity feature information after the entities in the first item and the second item are aligned, determining the item to be displayed according to the similarity calculation result, and displaying the item to be displayed.
  • the display module is also used to:
  • Item similarity is calculated according to the first similarity of each entity feature information after entity feature alignment.
  • an electronic device including:
  • processors one or more processors
  • the one or more processors When the one or more programs are executed by one or more processors, the one or more processors are caused to implement any of the above-described methods of displaying items.
  • a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, implements any one of the above-mentioned object display methods.
  • FIG. 1 is a schematic diagram of the main flow of an article display method provided according to a first embodiment of the present disclosure
  • Fig. 2a is a schematic diagram of the main flow of an article display method provided according to a second embodiment of the present disclosure
  • Fig. 2b is a schematic diagram of the main process of obtaining entity feature information corresponding to the first item in the method described in Fig. 2a;
  • Fig. 2c is a schematic diagram of the main process of calculating the similarity of items in the method described in Fig. 2a;
  • Fig. 2d is a schematic diagram showing the items to be displayed in the method described in Fig. 2a;
  • FIG. 3 is a schematic diagram of the main modules of the article display device provided according to an embodiment of the present disclosure.
  • FIG. 4 is an exemplary system architecture diagram to which embodiments of the present disclosure may be applied;
  • FIG. 5 is a schematic structural diagram of a computer system suitable for implementing a terminal device or a server according to an embodiment of the present disclosure.
  • FIG. 1 is a schematic diagram of the main process of the article display method provided according to the first embodiment of the present disclosure; as shown in FIG. 1 , the article display method provided by the embodiment of the present disclosure mainly includes:
  • Step S101 Obtain item information corresponding to the first item, perform entity feature extraction on the item information corresponding to the first item, and obtain entity feature information corresponding to the first item; wherein the item information includes item type and item name.
  • the item information can be obtained from the description information (such as title, introduction) of the item on the e-commerce platform. Since there may be multiple different descriptions for the same attribute, therefore, according to the present disclosure
  • the entity extraction technology of the knowledge graph can be used to extract the entity feature of the item information corresponding to the first item, and obtain the entity feature vector, so as to facilitate the subsequent determination of the item to be displayed and improve the accuracy of the determined displayed item. degree and determine the efficiency of displayed items.
  • Named entity extraction belongs to knowledge extraction, that is, knowledge extraction is performed from data of different sources and structures to form knowledge (structured data) and store it in the knowledge graph.
  • the above-mentioned item information further includes an item picture
  • the above-mentioned method further includes:
  • Image recognition processing is performed on the image of the item to obtain the corresponding text information and item attribute information.
  • the existing The convolutional neural network and other technologies perform image recognition processing on the picture, so as to obtain the corresponding text information and item attribute information, and then improve the item feature information of the first item, so as to facilitate the subsequent determination according to the item feature information of the first item. Display items to enhance user experience.
  • Step S102 determining the second item from the historical item set according to the entity feature information corresponding to the first item, and performing entity feature alignment on the entity feature information of the first item and the second item; wherein, the historical item set includes multiple items. entity feature information.
  • entity feature alignment mainly uses machine learning models of supervised learning, such as decision trees, support vector machines, ensemble learning, etc., and is implemented by attribute similarity matching. Relying on the attribute information of entity features, through attribute similarity, the inference of the alignment relationship of different entity features is carried out. Due to the different types of attributes, different attribute similarity calculation functions need to be designed, and different attribute similarity functions need to be designed for different fields. It should be noted that the above description about the entity feature alignment is only an example, and any existing entity feature alignment method can also be used.
  • the above method further includes a step of constructing the collection of historical items, including:
  • a collection of historical items is constructed based on entity feature information corresponding to multiple items.
  • the item information can be obtained from the same platform or from across platforms
  • a collection of historical items can be constructed, which is conducive to expanding the scope of item information and further improving the comprehensiveness of the determined items to be displayed. , improve the matching degree, and improve the user experience.
  • the first weight coefficient corresponding to the entity feature information is configured; the above-mentioned step of determining the second item from the historical item set according to the entity feature information corresponding to the first item includes:
  • the second item is determined from the historical item set according to the first weight coefficient of the entity feature information corresponding to the first item and the second item quantity threshold.
  • the second item can be determined according to the weight coefficient corresponding to the entity feature information, so as to reduce the amount of calculation for the subsequent calculation of the similarity of the items, reduce the calculation cost, and improve the efficiency of determining the item to be displayed.
  • the quantity of the second item may be one or more.
  • the entity feature information of the first item and the respective items in the historical item collection can also be directly matched.
  • the entity feature information is used to determine the items to be displayed by subsequently calculating the similarity of the items.
  • Step S103 Calculate the similarity of the items according to the entity feature information of the first item and the second item after the entity features are aligned, determine the item to be displayed according to the similarity calculation result, and display the item to be displayed.
  • the similarity of the items is calculated after the entity feature information of the first item and the second item is aligned, which is beneficial to reduce the calculation cost, improve the efficiency of determining the items to be displayed, and improve the matching accuracy of the displayed items. .
  • the step of determining the items to be displayed according to the similarity calculation result may be determined according to a similarity threshold and/or a threshold of the number of items to be displayed.
  • the above step of calculating the similarity of the items according to the entity feature information after the entity features in the first item and the second item are aligned includes:
  • Item similarity is calculated according to the first similarity of the entity feature information after entity alignment.
  • the second weight coefficient corresponding to each entity feature information after entity alignment is configured; the above-mentioned step of calculating the item similarity according to the first similarity of each entity feature information after entity feature alignment is further include:
  • Item similarity is calculated according to the first similarity of each entity feature information after entity feature alignment and the second weight coefficient corresponding to each entity feature information.
  • the corresponding second weight coefficients can be configured for each entity feature information according to user needs, etc., and then the similarity of items can be calculated according to the first similarity of each entity feature information and the second weight coefficient corresponding to each entity feature information.
  • the accuracy and efficiency of the displayed items are further improved, and the matching degree is improved.
  • the first weight coefficient and the second weight coefficient corresponding to the entity feature information may be the same, or may be configured separately according to actual user requirements.
  • the item information corresponding to the first item is acquired, entity feature extraction is performed on the item information corresponding to the first item, and the entity feature information corresponding to the first item is obtained; wherein the item information includes the item type and item name; determine the second item from the historical item collection according to the entity feature information corresponding to the first item, and perform entity feature alignment on the entity feature information of the first item and the second item; wherein, the historical item collection includes multiple items According to the entity feature information of the first item and the second item after the entity features are aligned, the similarity of the items is calculated, the items to be displayed are determined according to the similarity calculation results, and the technical means of displaying the items to be displayed are overcome, so the In the existing method for displaying items, the accuracy of the determined items to be displayed is low, the matching degree between the displayed items and the target items is low, the display efficiency is low, the calculation cost is high, and the user experience is poor. The accuracy and display efficiency of the items improve the matching degree, reduce the calculation cost
  • Fig. 2a is a schematic diagram of the main flow of the article display method provided according to the second embodiment of the present disclosure; as shown in Fig. 2a, the article display method provided by the embodiment of the present disclosure mainly includes:
  • Step S201 Obtain item information of multiple items, perform entity feature extraction on the item information respectively, and determine the entity feature information corresponding to each item; construct a historical item set based on the entity feature information corresponding to the multiple items.
  • the item information can be obtained from the same platform or from across platforms
  • a collection of historical items can be constructed, which is conducive to expanding the scope of item information and further improving the comprehensiveness of the determined items to be displayed. , improve the matching degree, and improve the user experience.
  • Step S202 Obtain item information corresponding to the first item, where the item information includes the item type, the item name, and the item picture.
  • item information may be obtained from description information (eg, title, introduction) of the item on the e-commerce platform.
  • step S203 image recognition processing is performed on the image of the item to obtain corresponding text information and item attribute information.
  • the existing The convolutional neural network and other technologies perform image recognition processing on the picture, so as to obtain the corresponding text information and item attribute information, and then improve the item feature information of the first item, so as to facilitate the subsequent determination according to the item feature information of the first item. Display items to enhance user experience.
  • Step S204 performing entity feature extraction on the item information corresponding to the first item to obtain entity feature information corresponding to the first item.
  • the entity extraction technology of the knowledge graph can be used to extract the entity feature of the item information corresponding to the first item, and obtain Entity feature vector, so as to facilitate the subsequent determination of the item to be displayed, improve the accuracy of the determined displayed item and the efficiency of determining the displayed item.
  • entity features are extracted from heterogeneous data and multimodal data (ie, different ways of presenting item information). Item information exists in the form of pictures, title descriptions, etc. It is necessary to use image recognition and pattern recognition to obtain text information in pictures, and to obtain the product (the first item) through technologies such as entity recognition, relationship extraction, and multimodal information extraction. ) feature information.
  • Step S205 configure the first weight coefficient corresponding to the entity feature information; determine the second item from the historical item set according to the first weight coefficient of the entity feature information corresponding to the first item and the second item quantity threshold.
  • the second item can be determined according to the weight coefficient corresponding to the entity feature information, so as to reduce the amount of calculation for the subsequent calculation of the similarity of the items, reduce the calculation cost, and improve the efficiency of determining the item to be displayed.
  • the quantity of the second item may be one or more.
  • Step S206 performing entity feature alignment on the entity feature information of the first item and the second item.
  • entity feature alignment is mainly realized by using a machine learning model of supervised learning, such as decision tree, support vector machine, ensemble learning, etc., through attribute similarity matching. Relying on the attribute information of entity features, through attribute similarity, the inference of the alignment relationship of different entity features is carried out. Due to the different types of attributes, different attribute similarity calculation functions need to be designed, and different attribute similarity functions need to be designed for different fields. It should be noted that the above description about the entity feature alignment is only an example, and any existing entity feature alignment method can also be used.
  • Step S207 Determine the entity feature information after the entity feature alignment in the first item and the second item, and calculate the first similarity corresponding to each entity feature information respectively.
  • Step S208 configure the second weight coefficient corresponding to each entity feature information after entity alignment; calculate the item similarity according to the first similarity of each entity feature information after entity feature alignment and the second weight coefficient corresponding to each entity feature information .
  • the corresponding second weight coefficients can be configured for each entity feature information according to user requirements, etc., and then the similarity of items can be calculated according to the first similarity of each entity feature information and the second weight coefficient corresponding to each entity feature information.
  • the accuracy and efficiency of the displayed items are further improved, and the matching degree is improved.
  • Figure 2c after obtaining enough features (including title, attribute, entity feature information in the picture), it is enough to represent the comprehensive information of the product (item), and then through the entity feature alignment technology, for Perform interactive feature extraction for each aligned entity feature information, calculate the similarity of each entity feature information in the first item and the second item (item 1 and item 2 in Figure 2c), and then calculate the item similarity (ie, the matching score).
  • the matching score can be calculated by calculating the average value of the similarity of the feature information of each entity; it is also possible to configure the weight coefficients of each attribute of different categories of commodities with the help of the rule engine, and combine the weight coefficients to determine the final output matching score.
  • Step S209 Determine the item to be displayed according to the similarity calculation result, and display the item to be displayed.
  • the similarity of the items is calculated after the entity feature information of the first item and the second item is aligned, which is beneficial to reduce the calculation cost, improve the efficiency of determining the items to be displayed, and improve the matching accuracy of the displayed items.
  • the step of determining the items to be displayed according to the similarity calculation result may be determined according to a similarity threshold and/or a threshold of the number of items to be displayed.
  • the information of the first item and the item to be displayed can be displayed side by side, so as to provide users with information of similar/same items more intuitively, so as to provide users with The choice of , provides more intuitive guidance and enhances the user experience.
  • the item information corresponding to the first item is acquired, entity feature extraction is performed on the item information corresponding to the first item, and the entity feature information corresponding to the first item is obtained; wherein the item information includes the item type and item name; determine the second item from the historical item collection according to the entity feature information corresponding to the first item, and perform entity feature alignment on the entity feature information of the first item and the second item; wherein, the historical item collection includes multiple items According to the entity feature information of the first item and the second item after the entity features are aligned, the similarity of the items is calculated, the items to be displayed are determined according to the similarity calculation results, and the technical means of displaying the items to be displayed are overcome, so the In the existing method for displaying items, the accuracy of the determined items to be displayed is low, the matching degree between the displayed items and the target items is low, the display efficiency is low, the calculation cost is high, and the user experience is poor. The accuracy and display efficiency of the items improve the matching degree, reduce the calculation cost
  • FIG. 3 is a schematic diagram of the main modules of the article display device provided according to the embodiment of the present disclosure; as shown in FIG. 3 , the article display device 300 provided by the embodiment of the present disclosure mainly includes:
  • the information acquisition module 301 is configured to acquire item information corresponding to the first item, perform entity feature extraction on the item information corresponding to the first item, and obtain entity feature information corresponding to the first item; wherein the item information includes item type and item name.
  • the item information can be obtained from the description information (such as title, introduction) of the item on the e-commerce platform. Since there may be multiple different descriptions for the same attribute, therefore, according to the present disclosure
  • the entity extraction technology of the knowledge graph can be used to extract the entity feature of the item information corresponding to the first item, and obtain the entity feature vector, so as to facilitate the subsequent determination of the item to be displayed and improve the displayed item. Accuracy and efficiency in determining displayed items.
  • Named entity extraction belongs to knowledge extraction, that is, knowledge extraction is performed from data of different sources and structures to form knowledge (structured data) and store it in the knowledge graph.
  • the above-mentioned item information further includes an item picture
  • the above-mentioned item display device 300 further includes a picture recognition processing module, and before the step of performing entity feature extraction on the item information corresponding to the first item, the picture recognition processing module Used for:
  • Image recognition processing is performed on the image of the item to obtain the corresponding text information and item attribute information.
  • the existing The convolutional neural network and other technologies perform image recognition processing on the picture, so as to obtain the corresponding text information and item attribute information, and then improve the item feature information of the first item, so as to facilitate the subsequent determination according to the item feature information of the first item. Display items to enhance user experience.
  • the entity feature alignment module 302 is configured to determine the second item from the historical item set according to the entity feature information corresponding to the first item, and perform entity feature alignment on the entity feature information of the first item and the second item; wherein, in the historical item set Contains entity feature information for multiple items.
  • entity feature alignment is mainly realized by using a machine learning model of supervised learning, such as decision tree, support vector machine, ensemble learning, etc., through attribute similarity matching. Relying on the attribute information of entity features, through attribute similarity, the inference of the alignment relationship of different entity features is carried out. Due to the different types of attributes, different attribute similarity calculation functions need to be designed, and different attribute similarity functions need to be designed for different fields. It should be noted that the above description about the entity feature alignment is only an example, and any existing entity feature alignment method can also be used.
  • the above-mentioned item display device 300 further includes a historical item collection building module, before the step of determining the second item from the historical item collection according to the entity feature information corresponding to the first item, the historical item collection building module. Used for:
  • a collection of historical items is constructed based on entity feature information corresponding to multiple items.
  • the item information can be obtained from the same platform or from across platforms
  • a collection of historical items can be constructed, which is conducive to expanding the scope of item information and further improving the comprehensiveness of the determined items to be displayed. , improve the matching degree, and improve the user experience.
  • the aforementioned entity feature alignment module 302 is further configured to:
  • the second item is determined from the historical item set according to the first weight coefficient of the entity feature information corresponding to the first item and the second item quantity threshold.
  • the second item can be determined according to the weight coefficient corresponding to the entity feature information, so as to reduce the amount of calculation for the subsequent calculation of the similarity of the items, reduce the calculation cost, and improve the efficiency of determining the item to be displayed.
  • the quantity of the second item may be one or more.
  • the entity feature information of the first item and the respective items in the historical item collection can also be directly matched.
  • the entity feature information is used to determine the items to be displayed by subsequently calculating the similarity of the items.
  • the display module 303 is configured to calculate the similarity of the items according to the entity feature information after the entities in the first item and the second item are aligned, determine the item to be displayed according to the similarity calculation result, and display the item to be displayed.
  • the similarity of the items is calculated, which is beneficial to reduce the calculation cost, improve the efficiency of determining the items to be displayed, and improve the matching accuracy of the displayed items.
  • the step of determining the items to be displayed according to the similarity calculation result may be determined according to a similarity threshold and/or a threshold of the number of items to be displayed.
  • the above-mentioned display module 303 is also used for:
  • Item similarity is calculated according to the first similarity of the entity feature information after entity alignment.
  • the second weight coefficient corresponding to each entity feature information after entity alignment is configured; the above-mentioned display module 303 is further configured to:
  • Item similarity is calculated according to the first similarity of each entity feature information after entity feature alignment and the second weight coefficient corresponding to each entity feature information.
  • the corresponding second weight coefficients can be configured for each entity feature information according to user needs, etc., and then the similarity of items can be calculated according to the first similarity of each entity feature information and the second weight coefficient corresponding to each entity feature information.
  • the accuracy and efficiency of the displayed items are further improved, and the matching degree is improved.
  • the item information corresponding to the first item is acquired, the entity feature extraction is performed on the item information corresponding to the first item, and the entity feature information corresponding to the first item is obtained; wherein, the item information includes the item type and item name; determine the second item from the historical item collection according to the entity feature information corresponding to the first item, and perform entity feature alignment on the entity feature information of the first item and the second item; wherein, the historical item collection includes multiple items According to the entity feature information of the first item and the second item after the entity features are aligned, the similarity of the items is calculated, and the items to be displayed are determined according to the similarity calculation result, and the technical means of displaying the items to be displayed are overcome.
  • the accuracy of the determined items to be displayed is low, the matching degree between the displayed items and the target items is low, the display efficiency is low, the calculation cost is high, and the user experience is poor.
  • the accuracy and display efficiency of the items improve the matching degree, reduce the calculation cost, and improve the technical effect of the user experience.
  • FIG. 4 illustrates an exemplary system architecture 400 to which an article display method or article display apparatus (adjusted according to a specific case) may be applied with embodiments of the present disclosure.
  • the system architecture 400 may include terminal devices 401, 402, 403, a network 404 and a server 405 (this architecture is only an example, and the components included in the specific architecture can be adjusted according to the specific application).
  • the network 404 is a medium used to provide a communication link between the terminal devices 401 , 402 , 403 and the server 405 .
  • the network 404 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.
  • the user can use the terminal devices 401, 402, 403 to interact with the server 405 through the network 404 to receive or send messages and the like.
  • Various communication client applications may be installed on the terminal devices 401 , 402 and 403 , such as shopping applications, web browser applications, item display applications, data processing applications, social platform software, etc. (only examples).
  • the terminal devices 401, 402, 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.
  • the server 405 may be a server that provides various services, such as a server (only an example) for users to use the terminal devices 401 , 402 , and 403 (for item display/data processing).
  • the server can analyze and process the received item information and other data, and feed back the processing results (for example, the second item, the item to be displayed—just an example) to the terminal device.
  • the article display method provided by the embodiment of the present disclosure is generally performed by the server 405 , and accordingly, the article display device is generally set in the server 405 .
  • terminal devices, networks and servers in FIG. 4 are merely illustrative. There can be any number of terminal devices, networks and servers according to implementation needs.
  • FIG. 5 shows a schematic structural diagram of a computer system 500 suitable for implementing a terminal device or a server according to an embodiment of the present disclosure.
  • the terminal device or server shown in FIG. 5 is only an example, and should not impose any limitation on the functions and scope of use of the embodiments of the present disclosure.
  • a computer system 500 includes a central processing unit (CPU) 501 which can be loaded into a random access memory (RAM) 503 according to a program stored in a read only memory (ROM) 502 or a program from a storage section 508 Instead, various appropriate actions and processes are performed.
  • RAM random access memory
  • ROM read only memory
  • various programs and data required for the operation of the system 500 are also stored.
  • the CPU 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504.
  • An input/output (I/O) interface 505 is also connected to bus 504 .
  • the following components are connected to the I/O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc. ; and a communication section 509 including a network interface card such as a LAN card, a modem, and the like. The communication section 509 performs communication processing via a network such as the Internet.
  • a drive 510 is also connected to the I/O interface 505 as needed.
  • a removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 510 as needed so that a computer program read therefrom is installed into the storage section 508 as needed.
  • embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the method illustrated in the flowchart.
  • the computer program may be downloaded and installed from the network via the communication portion 509 and/or installed from the removable medium 511 .
  • CPU central processing unit
  • the above-described functions defined in the system of the present disclosure are executed.
  • the computer-readable medium shown in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two.
  • the computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples of computer readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable Programmable read only memory (EPROM or flash memory), fiber optics, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
  • a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
  • a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code therein. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device .
  • Program code embodied on a computer readable medium may be transmitted using any suitable medium including, but not limited to, wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
  • each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more logical functions for implementing the specified functions executable instructions.
  • the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
  • the modules involved in the embodiments of the present disclosure may be implemented in software or hardware.
  • the described modules can also be set in the processor, for example, it can be described as: a processor includes an information acquisition module, an entity feature alignment module and a display module.
  • the names of these modules do not constitute a limitation of the module itself under certain circumstances.
  • the information acquisition module can also be described as "used to obtain the item information corresponding to the first item, and the item corresponding to the first item. information to perform entity feature extraction, and obtain the module of entity feature information corresponding to the first item".
  • the present disclosure also provides a computer-readable medium.
  • the computer-readable medium may be included in the device described in the above-mentioned embodiments, or it may exist alone without being assembled into the device.
  • the above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed by one of the equipment, the equipment includes: acquiring the item information corresponding to the first item, and performing the processing on the item information corresponding to the first item. Entity feature extraction to obtain entity feature information corresponding to the first item; wherein, the item information includes the item type and item name; the second item is determined from the historical item collection according to the entity feature information corresponding to the first item, and the first item and the first item are compared.
  • Entity feature alignment is performed on the entity feature information of the two items; wherein, the historical item set includes the entity feature information of multiple items; according to the entity feature information after the entity feature alignment in the first item and the second item, the similarity of the items is calculated, according to The similarity calculation result determines the item to be displayed, and displays the item to be displayed.
  • the item information corresponding to the first item is acquired, the entity feature extraction is performed on the item information corresponding to the first item, and the entity feature information corresponding to the first item is obtained; wherein, the item information includes the item type and item name; determine the second item from the historical item collection according to the entity feature information corresponding to the first item, and perform entity feature alignment on the entity feature information of the first item and the second item; wherein, the historical item collection includes multiple items According to the entity feature information of the first item and the second item after the entity features are aligned, the similarity of the items is calculated, the items to be displayed are determined according to the similarity calculation results, and the technical means of displaying the items to be displayed are overcome, so the In the existing method for displaying items, the accuracy of the determined items to be displayed is low, the matching degree between the displayed items and the target items is low, the display efficiency is low, the calculation cost is high, and the user experience is poor. The accuracy and display efficiency of the items improve the matching degree, reduce the

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Finance (AREA)
  • Artificial Intelligence (AREA)
  • Accounting & Taxation (AREA)
  • Evolutionary Computation (AREA)
  • Multimedia (AREA)
  • Evolutionary Biology (AREA)
  • General Business, Economics & Management (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

一种物品展示方法和装置，方法包括：获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息（S101）；其中，物品信息包括物品类型和物品名称；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐（S102）；其中，历史物品集合中包括多个物品的实体特征信息；根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品（S103）。

Description

物品展示方法和装置
本申请要求于2021年1月29日提交的题为“一种物品展示方法和装置”的中国专利申请No.202110130625.X的优先权，在此全文引用上述中国专利申请公开的内容以作为本申请的全部或部分内容。
技术领域
本公开涉及计算机技术领域，尤其涉及物品展示方法和装置。
背景技术
随着网购的兴起，在当前的B2B(Business-to-Business,企业对企业)和B2C(Business-to-Consumer，企业对用户)电商平台上，物品的种类、数量呈爆发式增长。如何在用户在购买物品时提供相应的同品比较、替代品与相似品推荐场景，即向用户展示属性相同或类似的物品，以供用户进行选择，有利于提升用户体验，增加用户粘性。
现有技术中至少存在如下问题：
现有的物品展示方法中，存在所确定的展示物品的准确率较低、展示物品与目标物品的匹配度较低，展示效率低、计算成本较高，用户体验差的技术问题。
发明内容
有鉴于此，本公开实施例提供一种物品展示方法和装置，能够提高展示物品的准确率和展示效率，提升了匹配度，降低计算成本，提升用户体验。
为实现上述目的，根据本公开实施例的第一方面，提供了一种物品展示方法，包括：
获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息 包括物品类型和物品名称；
根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；
根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品。
进一步地，物品信息还包括物品图片，在对第一物品对应的物品信息进行实体特征提取的步骤之前，方法还包括：
对物品图片进行图片识别处理，以获取相应的文本信息和物品属性信息。
进一步地，在根据第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤之前，方法还包括历史物品集合的构建步骤，包括：
获取多个物品的物品信息，分别对物品信息进行实体特征提取，确定各物品对应的实体特征信息；
基于多个物品对应的实体特征信息构建历史物品集合。
进一步地，配置实体特征信息对应的第一权重系数；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤包括：
根据第一物品对应的实体特征信息的第一权重系数、第二物品数量阈值从历史物品集合中确定第二物品。
进一步地，根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度的步骤包括：
确定第一物品和第二物品中实体特征对齐后的实体特征信息，分别计算各实体特征信息对应的第一相似度；
根据实体对齐后的各实体特征信息的第一相似度，计算物品相似 度。
进一步地，配置实体对齐后的各实体特征信息对应的第二权重系数；根据实体特征对齐后的各实体特征信息的第一相似度，计算物品相似度的步骤，还包括：
根据实体特征对齐后的各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，计算物品相似度。
根据本公开实施例的第二方面，提供了一种物品展示装置，包括：
信息获取模块，用于获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称；
实体特征对齐模块，用于根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；
展示模块，用于根据第一物品和第二物品中实体对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品。
进一步地，展示模块还用于：
确定第一物品和第二物品中实体特征对齐后的各实体特征信息；
依次计算各实体特征信息对应的第一相似度；
根据实体特征对齐后的各实体特征信息的第一相似度，计算物品相似度。
根据本公开实施例的第三方面，提供了一种电子设备，包括：
一个或多个处理器；
存储装置，用于存储一个或多个程序，
当一个或多个程序被一个或多个处理器执行，使得一个或多个处 理器实现如上述任一种物品展示方法。
根据本公开实施例的第四方面，提供了一种计算机可读介质，其上存储有计算机程序，该程序被处理器执行时实现如上述任一种物品展示方法。
上述的非惯用的可选方式所具有的进一步效果将在下文中结合具体实施方式加以说明。
附图说明
附图用于更好地理解本公开，不构成对本公开的不当限定。其中：
图1是根据本公开第一实施例提供的物品展示方法的主要流程的示意图；
图2a是根据本公开第二实施例提供的物品展示方法的主要流程的示意图；
图2b是图2a所述方法中得到第一物品对应的实体特征信息的主要流程的示意图；
图2c是图2a所述方法中计算物品相似度的主要流程的示意图；
图2d是图2a所述方法中展示待展示物品的示意图；
图3是根据本公开实施例提供的物品展示装置的主要模块的示意图；
图4是本公开实施例可以应用于其中的示例性系统架构图；
图5是适于用来实现本公开实施例的终端设备或服务器的计算机系统的结构示意图。
具体实施方式
以下结合附图对本公开的示范性实施例做出说明，其中包括本公开实施例的各种细节以助于理解，应当将它们认为仅仅是示范性的。因此，本领域普通技术人员应当认识到，可以对这里描述的实施例做出各种改变和修改，而不会背离本公开的范围和精神。同样，为了清 楚和简明，以下的描述中省略了对公知功能和结构的描述。
图1是根据本公开第一实施例提供的物品展示方法的主要流程的示意图；如图1所示，本公开实施例提供的物品展示方法主要包括：
步骤S101，获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称。
具体地，根据本公开实施例，物品信息可以从电商平台对于物品的描述信息(如标题、介绍)中获取，由于对于相同属性的特征可能存在多种不同的描述，因此，根据本公开实施例的一具体实施方式，可以采取知识图谱的实体提取技术来对第一物品对应的物品信息进行实体特征提取，得到实体特征向量，以便于后续确定待展示物品，提升所确定的展示物品的准确度和确定展示物品的效率。
其中，实体特征抽取：即命名实体抽取，包括实体的检测和分类，属于知识抽取，即从不同来源、不同结构的数据中进行知识抽取，形成知识(结构化数据)存入至知识图谱中。
进一步地，根据本公开实施例，上述物品信息还包括物品图片，在对第一物品对应的物品信息进行实体特征提取的步骤之前，上述方法还包括：
对物品图片进行图片识别处理，以获取相应的文本信息和物品属性信息。
通过上述设置，对于电商平台中售卖的物品而言，较多情况下，会同时采取图像形式来展示物品详情，为了充分提取物品的实体特征信息，根据本公开实施例，可以采用现有的卷积神经网络等技术对图片进行图片识别处理，以从获取相应的文本信息和物品属性信息，进而完善第一物品的物品特征信息，以便于后续根据该第一物品的物品特征信息来确定待展示物品，提升用户体验。
步骤S102，根据第一物品对应的实体特征信息从历史物品集合中 确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息。
具体地，根据本公开实施例，实体特征对齐：主要是利用监督学习的机器学习模型，如决策树、支持向量机、集成学习等，是通过属性相似度匹配的方式实现。依赖实体特征的属性信息，通过属性相似度，进行不同实体特征对齐关系的推断。由于属性的类别不同，需要设计不同的属性相似度计算函数，且不同的领域需要设计不同的属性相似度函数。需要说明的是，上述关于实体特征对齐的描述仅为示例，还可以采用现有的任意一种实体特征对齐方式实现。
进一步地，根据本公开实施例，在根据第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤之前，上述方法还包括历史物品集合的构建步骤，包括：
获取多个物品的物品信息，分别对物品信息进行实体特征提取，确定各物品对应的实体特征信息；
基于多个物品对应的实体特征信息构建历史物品集合。
通过上述设置，基于多个物品信息(该物品信息可以从同一平台获取，也可以跨平台获取)，进而构建历史物品集合，有利于拓展物品信息的范围，进一步提升所确定的待展示物品的全面性，提升匹配度，提升用户体验。
优选地，根据本公开实施例，配置实体特征信息对应的第一权重系数；上述根据第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤包括：
根据第一物品对应的实体特征信息的第一权重系数、第二物品数量阈值从历史物品集合中确定第二物品。
通过上述设置，可以根据实体特征信息对应的权重系数来确定第二物品，以减少后续计算物品相似度的计算量，降低计算成本，提升确定待展示物品的效率。其中，第二物品的数量可以为一个或多个。
根据本公开实施例的一具体实施方式，若根据第一物品对应的实体特征信息未匹配到相应的第二物品，也可直接根据第一物品的实体特征信息与历史物品集合中的各物品的实体特征信息，通过后续计算物品相似度的方式，确定待展示物品。
步骤S103，根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品。
通过上述设置，通过将第一物品与第二物品的实体特征信息进行实体特征对齐之后，再计算物品相似度，有利于降低计算成本，提升确定待展示物品的效率，提升展示物品的匹配准确率。
其中，根据相似度计算结果确定待展示物品的步骤，可以根据相似度阈值和/或待展示物品数量阈值进行确定。
进一步地，根据本公开实施例，上述根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度的步骤包括：
确定第一物品和第二物品中实体特征对齐后的实体特征信息，分别计算各实体特征信息对应的第一相似度；
根据实体对齐后的各实体特征信息的第一相似度，计算物品相似度。
通过上述设置，分别计算实体特征对齐后的各个实体特征信息的相似度，进而根据各实体特征信息的相似度计算物品相似度，更有利里为用户展示与第一物品全面匹配的待展示物品，提升用户体验。
优选地，根据本公开实施例，配置实体对齐后的各实体特征信息对应的第二权重系数；上述根据实体特征对齐后的各实体特征信息的第一相似度，计算物品相似度的步骤，还包括：
根据实体特征对齐后的各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，计算物品相似度。
通过上述设置，可根据用户需求等来为各实体特征信息配置相应 的第二权重系数，进而根据各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，来计算物品相似度，进一步提高了展示物品的准确率和展示效率，提升了匹配度。
根据本公开实施例的另一具体实施方式，上述实体特征信息对应的第一权重系数和第二权重系数可以一致，也可以根据实际用户需求分别配置。
根据本公开实施例的技术方案，因为采用获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品的技术手段，所以克服了现有的物品展示方法中，所确定的展示物品的准确率较低、展示物品与目标物品的匹配度较低，展示效率低、计算成本较高，用户体验差的技术问题，进而达到提高展示物品的准确率和展示效率，提升了匹配度，降低计算成本，提升用户体验的技术效果。
图2a是根据本公开第二实施例提供的物品展示方法的主要流程的示意图；如图2a所示，本公开实施例提供的物品展示方法主要包括：
步骤S201，获取多个物品的物品信息，分别对物品信息进行实体特征提取，确定各物品对应的实体特征信息；基于多个物品对应的实体特征信息构建历史物品集合。
通过上述设置，基于多个物品信息(该物品信息可以从同一平台获取，也可以跨平台获取)，进而构建历史物品集合，有利于拓展物品信息的范围，进一步提升所确定的待展示物品的全面性，提升匹配度，提升用户体验。
步骤S202，获取第一物品对应的物品信息，其中，上述物品信息包括物品类型、物品名称和物品图片。
具体地，根据本公开实施例，物品信息可以从电商平台对于物品的描述信息(如标题、介绍)中获取。
步骤S203，对物品图片进行图片识别处理，以获取相应的文本信息和物品属性信息。
通过上述设置，对于电商平台中售卖的物品而言，较多情况下，会同时采取图像形式来展示物品详情，为了充分提取物品的实体特征信息，根据本公开实施例，可以采用现有的卷积神经网络等技术对图片进行图片识别处理，以从获取相应的文本信息和物品属性信息，进而完善第一物品的物品特征信息，以便于后续根据该第一物品的物品特征信息来确定待展示物品，提升用户体验。
步骤S204，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息。
由于对于相同属性的特征可能存在多种不同的描述，因此，根据本公开实施例的一具体实施方式，可以采取知识图谱的实体提取技术来对第一物品对应的物品信息进行实体特征提取，得到实体特征向量，以便于后续确定待展示物品，提升所确定的展示物品的准确度和确定展示物品的效率。具体地，如图2b所示，从多远异构数据、多模态数据(即不同物品信息呈现方式)中提取实体特征。物品信息以图片、标题描述等形式存在，需要使用图像识别和模式识别获取到图片中的文本信息，并通过实体识别、关系抽取、多模态信息抽取等技术，获取到该商品(第一物品)的特征信息。
步骤S205，配置实体特征信息对应的第一权重系数；根据第一物品对应的实体特征信息的第一权重系数、第二物品数量阈值从历史物品集合中确定第二物品。
通过上述设置，可以根据实体特征信息对应的权重系数来确定第 二物品，以减少后续计算物品相似度的计算量，降低计算成本，提升确定待展示物品的效率。其中，第二物品的数量可以为一个或多个。
步骤S206，将第一物品和第二物品的实体特征信息进行实体特征对齐。
具体地，根据本公开实施例，实体特征对齐，主要是利用监督学习的机器学习模型，如决策树、支持向量机、集成学习等，是通过属性相似度匹配的方式实现。依赖实体特征的属性信息，通过属性相似度，进行不同实体特征对齐关系的推断。由于属性的类别不同，需要设计不同的属性相似度计算函数，且不同的领域需要设计不同的属性相似度函数。需要说明的是，上述关于实体特征对齐的描述仅为示例，还可以采用现有的任意一种实体特征对齐方式实现。
步骤S207，确定第一物品和第二物品中实体特征对齐后的实体特征信息，分别计算各实体特征信息对应的第一相似度。
通过上述设置，分别计算实体特征对齐后的各个实体特征信息的相似度，进而根据各实体特征信息的相似度计算物品相似度，更有利里为用户展示与第一物品全面匹配的待展示物品，提升用户体验。
步骤S208，配置实体对齐后的各实体特征信息对应的第二权重系数；根据实体特征对齐后的各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，计算物品相似度。
通过上述设置，可根据用户需求等来为各实体特征信息配置相应的第二权重系数，进而根据各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，来计算物品相似度，进一步提高了展示物品的准确率和展示效率，提升了匹配度。具体地，如图2c所示，获取到足够多的特征后(包括标题、属性、图片中的实体特征信息)，则足以表征这个商品(物品)的全面信息，然后通过实体特征对齐技术，针对每个对齐后的实体特征信息进行交互特征提取，计算第一物品和第二物品(图2c中的商品1和商品2)中各个实体特征信息的相 似度，进而计算物品相似度(即匹配得分)。匹配得分可以通过计算各个实体特征信息的相似度的平均值；也可以借助规则引擎对不同品类的商品各属性的权重系数进行配置，结合权重系数确定最终输出的匹配得分。
步骤S209，根据相似度计算结果确定待展示物品，并展示待展示物品。
通过上述设置，通过将第一物品与第二物品的实体特征信息进行实体特征对齐之后，再计算物品相似度，有利于降低计算成本，提升确定待展示物品的效率，提升展示物品的匹配准确率。其中，根据相似度计算结果确定待展示物品的步骤，可以根据相似度阈值和/或待展示物品数量阈值进行确定。
具体地，如图2d所示，展示上述确定的待展示物品时，可以将第一物品和待展示物品的信息进行并列展示，以更直观的为用户提供相似/相同物品的信息，以便为用户的选择提供较为直观的指导，提升用户体验。
根据本公开实施例的技术方案，因为采用获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品的技术手段，所以克服了现有的物品展示方法中，所确定的展示物品的准确率较低、展示物品与目标物品的匹配度较低，展示效率低、计算成本较高，用户体验差的技术问题，进而达到提高展示物品的准确率和展示效率，提升了匹配度，降低计算成本，提升用户体验的技术效果。
图3是根据本公开实施例提供的物品展示装置的主要模块的示意图；如图3所示，本公开实施例提供的物品展示装置300主要包括：
信息获取模块301，用于获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称。
具体地，根据本公开实施例，物品信息可以从电商平台对于物品的描述信息(如标题、介绍)中获取，由于对于相同属性的特征可能存在多种不同的描述，因此，根据本公开实施例的一具体时实施方式，可以采取知识图谱的实体提取技术来对第一物品对应的物品信息进行实体特征提取，得到实体特征向量，以便于后续确定待展示物品，提升所确定的展示物品的准确度和确定展示物品的效率。
其中，实体特征抽取：即命名实体抽取，包括实体的检测和分类，属于知识抽取，即从不同来源、不同结构的数据中进行知识抽取，形成知识(结构化数据)存入至知识图谱中。
进一步地，根据本公开实施例，上述物品信息还包括物品图片；上述物品展示装置300还包括图片识别处理模块，在对第一物品对应的物品信息进行实体特征提取的步骤之前，图片识别处理模块用于：
对物品图片进行图片识别处理，以获取相应的文本信息和物品属性信息。
通过上述设置，对于电商平台中售卖的物品而言，较多情况下，会同时采取图像形式来展示物品详情，为了充分提取物品的实体特征信息，根据本公开实施例，可以采用现有的卷积神经网络等技术对图片进行图片识别处理，以从获取相应的文本信息和物品属性信息，进而完善第一物品的物品特征信息，以便于后续根据该第一物品的物品特征信息来确定待展示物品，提升用户体验。
实体特征对齐模块302，用于根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特 征信息。
具体地，根据本公开实施例，实体特征对齐，主要是利用监督学习的机器学习模型，如决策树、支持向量机、集成学习等，是通过属性相似度匹配的方式实现。依赖实体特征的属性信息，通过属性相似度，进行不同实体特征对齐关系的推断。由于属性的类别不同，需要设计不同的属性相似度计算函数，且不同的领域需要设计不同的属性相似度函数。需要说明的是，上述关于实体特征对齐的描述仅为示例，还可以采用现有的任意一种实体特征对齐方式实现。
进一步地，根据本公开实施例，上述物品展示装置300还包括历史物品集合构建模块，在根据第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤之前，历史物品集合构建模块用于：
获取多个物品的物品信息，分别对物品信息进行实体特征提取，确定各物品对应的实体特征信息；
基于多个物品对应的实体特征信息构建历史物品集合。
通过上述设置，基于多个物品信息(该物品信息可以从同一平台获取，也可以跨平台获取)，进而构建历史物品集合，有利于拓展物品信息的范围，进一步提升所确定的待展示物品的全面性，提升匹配度，提升用户体验。
优选地，根据本公开实施例，配置实体特征信息对应的第一权重系数；上述实体特征对齐模块302还用于：
根据第一物品对应的实体特征信息的第一权重系数、第二物品数量阈值从历史物品集合中确定第二物品。
通过上述设置，可以根据实体特征信息对应的权重系数来确定第二物品，以减少后续计算物品相似度的计算量，降低计算成本，提升确定待展示物品的效率。其中，第二物品的数量可以为一个或多个。
根据本公开实施例的一具体实施方式，若根据第一物品对应的实体特征信息未匹配到相应的第二物品，也可直接根据第一物品的实体 特征信息与历史物品集合中的各物品的实体特征信息，通过后续计算物品相似度的方式，确定待展示物品。
展示模块303，用于根据第一物品和第二物品中实体对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品。
通过上述设置，通过将第一物品与第二物品的实体特征信息进行实体特征对齐之后，再计算物品相似度，有利于降低计算成本，提升确定待展示物品的效率，提升展示物品的匹配准确率。其中，根据相似度计算结果确定待展示物品的步骤，可以根据相似度阈值和/或待展示物品数量阈值进行确定。
进一步地，根据本公开实施例，上述展示模块303还用于：
确定第一物品和第二物品中实体特征对齐后的实体特征信息，分别计算各实体特征信息对应的第一相似度；
根据实体对齐后的各实体特征信息的第一相似度，计算物品相似度。
通过上述设置，分别计算实体特征对齐后的各个实体特征信息的相似度，进而根据各实体特征信息的相似度计算物品相似度，更有利里为用户展示与第一物品全面匹配的待展示物品，提升用户体验。
优选地，根据本公开实施例，配置实体对齐后的各实体特征信息对应的第二权重系数；上述展示模块303还用于：
根据实体特征对齐后的各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，计算物品相似度。
通过上述设置，可根据用户需求等来为各实体特征信息配置相应的第二权重系数，进而根据各实体特征信息的第一相似度和各实体特征信息对应的第二权重系数，来计算物品相似度，进一步提高了展示物品的准确率和展示效率，提升了匹配度。
根据本公开实施例的技术方案，因为采用获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品的技术手段，所以克服了现有的物品展示方法中，所确定的展示物品的准确率较低、展示物品与目标物品的匹配度较低，展示效率低、计算成本较高，用户体验差的技术问题，进而达到提高展示物品的准确率和展示效率，提升了匹配度，降低计算成本，提升用户体验的技术效果。
图4示出了可以应用本公开实施例的物品展示方法或物品展示装置(根据具体案件调整)的示例性系统架构400。
如图4所示，系统架构400可以包括终端设备401、402、403，网络404和服务器405(此架构仅仅是示例，具体架构中包含的组件可以根据申请具体情况调整)。网络404用以在终端设备401、402、403和服务器405之间提供通信链路的介质。网络404可以包括各种连接类型，例如有线、无线通信链路或者光纤电缆等等。
用户可以使用终端设备401、402、403通过网络404与服务器405交互，以接收或发送消息等。终端设备401、402、403上可以安装有各种通讯客户端应用，例如购物类应用、网页浏览器应用、物品展示类应用、数据处理类应用、社交平台软件等(仅为示例)。
终端设备401、402、403可以是具有显示屏并且支持网页浏览的各种电子设备，包括但不限于智能手机、平板电脑、膝上型便携计算机和台式计算机等等。
服务器405可以是提供各种服务的服务器，例如对用户利用终端设备401、402、403所(进行物品展示/进行数据处理)的服务器(仅为示例)。该服务器可以对接收到的物品信息等数据进行分析等处理，并将处理结果(例如第二物品、待展示物品--仅为示例)反馈给终端设备。
需要说明的是，本公开实施例所提供的物品展示方法一般由服务器405执行，相应地，物品展示装置一般设置于服务器405中。
应该理解，图4中的终端设备、网络和服务器的数目仅仅是示意性的。根据实现需要，可以具有任意数目的终端设备、网络和服务器。
下面参考图5，其示出了适于用来实现本公开实施例的终端设备或服务器的计算机系统500的结构示意图。图5示出的终端设备或服务器仅仅是一个示例，不应对本公开实施例的功能和使用范围带来任何限制。
如图5所示，计算机系统500包括中央处理单元(CPU)501，其可以根据存储在只读存储器(ROM)502中的程序或者从存储部分508加载到随机访问存储器(RAM)503中的程序而执行各种适当的动作和处理。在RAM 503中，还存储有系统500操作所需的各种程序和数据。CPU 501、ROM 502以及RAM 503通过总线504彼此相连。输入/输出(I/O)接口505也连接至总线504。
以下部件连接至I/O接口505：包括键盘、鼠标等的输入部分506；包括诸如阴极射线管(CRT)、液晶显示器(LCD)等以及扬声器等的输出部分507；包括硬盘等的存储部分508；以及包括诸如LAN卡、调制解调器等的网络接口卡的通信部分509。通信部分509经由诸如因特网的网络执行通信处理。驱动器510也根据需要连接至I/O接口505。 可拆卸介质511，诸如磁盘、光盘、磁光盘、半导体存储器等等，根据需要安装在驱动器510上，以便于从其上读出的计算机程序根据需要被安装入存储部分508。
特别地，根据本公开的实施例，上文参考流程图描述的过程可以被实现为计算机软件程序。例如，本公开的实施例包括一种计算机程序产品，其包括承载在计算机可读介质上的计算机程序，该计算机程序包含用于执行流程图所示的方法的程序代码。在这样的实施例中，该计算机程序可以通过通信部分509从网络上被下载和安装，和/或从可拆卸介质511被安装。在该计算机程序被中央处理单元(CPU)501执行时，执行本公开的系统中限定的上述功能。
需要说明的是，本公开所示的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子可以包括但不限于：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机访问存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本公开中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。而在本公开中，计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限 于：无线、电线、光缆、RF等等，或者上述的任意合适的组合。
附图中的流程图和框图，图示了按照本公开各种实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分，上述模块、程序段、或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。也应当注意，在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个接连地表示的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图或流程图中的每个方框、以及框图或流程图中的方框的组合，可以用执行规定的功能或操作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。
描述于本公开实施例中所涉及到的模块可以通过软件的方式实现，也可以通过硬件的方式来实现。所描述的模块也可以设置在处理器中，例如，可以描述为：一种处理器包括信息获取模块、实体特征对齐模块和展示模块。其中，这些模块的名称在某种情况下并不构成对该模块本身的限定，例如，信息获取模块还可以被描述为“用于获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息的模块”。
作为另一方面，本公开还提供了一种计算机可读介质，该计算机可读介质可以是上述实施例中描述的设备中所包含的；也可以是单独存在，而未装配入该设备中。上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被一个该设备执行时，使得该设备包括：获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息 进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品。
根据本公开实施例的技术方案，因为采用获取第一物品对应的物品信息，对第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，物品信息包括物品类型和物品名称；根据第一物品对应的实体特征信息从历史物品集合中确定第二物品，将第一物品和第二物品的实体特征信息进行实体特征对齐；其中，历史物品集合中包括多个物品的实体特征信息；根据第一物品和第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示待展示物品的技术手段，所以克服了现有的物品展示方法中，所确定的展示物品的准确率较低、展示物品与目标物品的匹配度较低，展示效率低、计算成本较高，用户体验差的技术问题，进而达到提高展示物品的准确率和展示效率，提升了匹配度，降低计算成本，提升用户体验的技术效果。
上述具体实施方式，并不构成对本公开保护范围的限制。本领域技术人员应该明白的是，取决于设计要求和其他因素，可以发生各种各样的修改、组合、子组合和替代。任何在本公开的精神和原则之内所作的修改、等同替换和改进等，均应包含在本公开保护范围之内。

Claims (10)

  1. 一种物品展示方法，包括：
    获取第一物品对应的物品信息，对所述第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，所述物品信息包括物品类型和物品名称；
    根据所述第一物品对应的实体特征信息从历史物品集合中确定第二物品，将所述第一物品和所述第二物品的实体特征信息进行实体特征对齐；其中，所述历史物品集合中包括多个物品的实体特征信息；
    根据所述第一物品和所述第二物品中实体特征对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示所述待展示物品。
  2. 根据权利要求1所述的物品展示方法，其中，所述物品信息还包括物品图片，在对所述第一物品对应的物品信息进行实体特征提取的步骤之前，所述方法还包括：
    对所述物品图片进行图片识别处理，以获取相应的文本信息和物品属性信息。
  3. 根据权利要求2所述的物品展示方法，其中，在所述根据所述第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤之前，所述方法还包括历史物品集合的构建步骤，包括：
    获取多个物品的物品信息，分别对所述物品信息进行实体特征提取，确定各物品对应的实体特征信息；
    基于所述多个物品对应的实体特征信息构建历史物品集合。
  4. 根据权利要求1所述的物品展示方法，其中，配置实体特征信息对应的第一权重系数；所述根据所述第一物品对应的实体特征信息从历史物品集合中确定第二物品的步骤包括：
    根据所述第一物品对应的实体特征信息的第一权重系数、第二物 品数量阈值从所述历史物品集合中确定第二物品。
  5. 根据权利要求1所述的物品展示方法，其中，所述根据所述第一物品和所述第二物品中实体特征对齐后的实体特征信息，计算物品相似度的步骤包括：
    确定所述第一物品和所述第二物品中实体特征对齐后的实体特征信息，分别计算各实体特征信息对应的第一相似度；
    根据实体对齐后的各实体特征信息的第一相似度，计算物品相似度。
  6. 根据权利要求5所述的物品展示方法，其中，配置实体对齐后的各实体特征信息对应的第二权重系数；所述根据实体特征对齐后的各实体特征信息的第一相似度，计算物品相似度的步骤，还包括：
    根据实体特征对齐后的各实体特征信息的第一相似度和所述各实体特征信息对应的第二权重系数，计算物品相似度。
  7. 一种物品展示装置，包括：
    信息获取模块，用于获取第一物品对应的物品信息，对所述第一物品对应的物品信息进行实体特征提取，得到第一物品对应的实体特征信息；其中，所述物品信息包括物品类型和物品名称；
    实体特征对齐模块，用于根据所述第一物品对应的实体特征信息从历史物品集合中确定第二物品，将所述第一物品和所述第二物品的实体特征信息进行实体特征对齐；其中，所述历史物品集合中包括多个物品的实体特征信息；
    展示模块，用于根据所述第一物品和所述第二物品中实体对齐后的实体特征信息，计算物品相似度，根据相似度计算结果确定待展示物品，并展示所述待展示物品。
  8. 根据权利要求7所述的物品展示装置，其中，所述展示模块还用于：
    确定所述第一物品和所述第二物品中实体特征对齐后的各实体特征信息；
    依次计算各实体特征信息对应的第一相似度；
    根据实体特征对齐后的各实体特征信息的第一相似度，计算物品相似度。
  9. 一种电子设备，包括：
    一个或多个处理器；
    存储装置，用于存储一个或多个程序，
    当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如权利要求1-6中任一所述的方法。
  10. 一种计算机可读介质，其上存储有计算机程序，所述程序被处理器执行时实现如权利要求1-6中任一所述的方法。
PCT/CN2022/070059 2021-01-29 2022-01-04 物品展示方法和装置 WO2022161116A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110130625.X 2021-01-29
CN202110130625.XA CN112861684A (zh) 2021-01-29 2021-01-29 一种物品展示方法和装置

Publications (1)

Publication Number Publication Date
WO2022161116A1 true WO2022161116A1 (zh) 2022-08-04

Family

ID=75987332

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/070059 WO2022161116A1 (zh) 2021-01-29 2022-01-04 物品展示方法和装置

Country Status (2)

Country Link
CN (1) CN112861684A (zh)
WO (1) WO2022161116A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112861684A (zh) * 2021-01-29 2021-05-28 北京电解智科技有限公司 一种物品展示方法和装置

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019123714A1 (ja) * 2017-12-21 2019-06-27 日本電気株式会社 情報処理装置、商品推薦方法、およびプログラム
CN110084658A (zh) * 2018-01-26 2019-08-02 北京京东尚科信息技术有限公司 物品匹配的方法和装置
CN110400201A (zh) * 2019-07-30 2019-11-01 中国工商银行股份有限公司 信息展示方法、装置、电子设备以及介质
CN110807095A (zh) * 2018-08-01 2020-02-18 北京京东尚科信息技术有限公司 一种物品匹配方法和装置
CN112036981A (zh) * 2020-09-02 2020-12-04 珠海随变科技有限公司 一种用于提供目标对比商品的方法、装置、设备及介质
CN112861684A (zh) * 2021-01-29 2021-05-28 北京电解智科技有限公司 一种物品展示方法和装置

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6564206B1 (en) * 1998-10-05 2003-05-13 Canon Kabushiki Kaisha Information search apparatus and method, and storage medium
CN103400286B (zh) * 2013-08-02 2016-12-28 世纪禾光科技发展（北京）有限公司 一种基于用户行为进行物品特征标注的推荐系统及方法
CN105953520B (zh) * 2016-05-06 2018-08-10 青岛海尔股份有限公司 智能冰箱控制方法及其控制系统
US11113659B2 (en) * 2016-08-19 2021-09-07 Stitch Fix, Inc. Systems and methods for improving recommendation systems
CN107861972B (zh) * 2017-09-15 2022-02-22 广州唯品会研究院有限公司 一种用户录入商品信息后显示商品全结果的方法及设备
KR102186077B1 (ko) * 2018-10-18 2020-12-04 카페24 주식회사 온라인 쇼핑몰 기반의 상품을 검색하는 방법, 상기 방법을 사용하는 장치 및 시스템
US11257028B2 (en) * 2019-01-07 2022-02-22 United Parcel Service Of America, Inc. System and methods for self-adjusting electronic reconciliation of a contribution amount and delivery value
CN110928894B (zh) * 2019-11-18 2023-05-02 北京秒针人工智能科技有限公司 实体对齐的方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019123714A1 (ja) * 2017-12-21 2019-06-27 日本電気株式会社 情報処理装置、商品推薦方法、およびプログラム
CN110084658A (zh) * 2018-01-26 2019-08-02 北京京东尚科信息技术有限公司 物品匹配的方法和装置
CN110807095A (zh) * 2018-08-01 2020-02-18 北京京东尚科信息技术有限公司 一种物品匹配方法和装置
CN110400201A (zh) * 2019-07-30 2019-11-01 中国工商银行股份有限公司 信息展示方法、装置、电子设备以及介质
CN112036981A (zh) * 2020-09-02 2020-12-04 珠海随变科技有限公司 一种用于提供目标对比商品的方法、装置、设备及介质
CN112861684A (zh) * 2021-01-29 2021-05-28 北京电解智科技有限公司 一种物品展示方法和装置

Also Published As

Publication number Publication date
CN112861684A (zh) 2021-05-28

Similar Documents

Publication Publication Date Title
WO2018192491A1 (zh) 信息推送方法和装置
US20140244388A1 (en) Social Content Synchronization
CN107870984A (zh) 识别搜索词的意图的方法和装置
CN109697641A (zh) 计算商品相似度的方法和装置
CN110020162B (zh) 用户识别方法和装置
CN110929136A (zh) 一种个性化推荐方法和装置
CN111160847B (zh) 一种处理流程信息的方法和装置
CN111815738B (zh) 一种构建地图的方法和装置
CN110827101B (zh) 一种店铺推荐的方法和装置
CN113450172B (zh) 一种商品推荐方法和装置
CN113495991A (zh) 一种推荐方法和装置
CN111160410B (zh) 一种物体检测方法和装置
WO2022161116A1 (zh) 物品展示方法和装置
CN111782850A (zh) 一种基于手绘图的物品搜索方法和装置
CN111988668B (zh) 一种视频推荐方法、装置、计算机设备及存储介质
CN108959289B (zh) 网站类别获取方法和装置
US20230085684A1 (en) Method of recommending data, electronic device, and medium
CN107679030B (zh) 基于用户操作行为数据提取同义词的方法和装置
CN113722593B (zh) 事件数据处理方法、装置、电子设备和介质
WO2022100401A1 (zh) 基于图像识别的价格信息处理方法、装置、设备和介质
CN113327145B (zh) 一种物品推荐方法和装置
CN110751496B (zh) 一种商品价格检测方法和装置
CN110110199B (zh) 信息输出方法和装置
CN112905786A (zh) 一种标签推荐方法和装置
CN111339124A (zh) 显示数据方法、装置、电子设备和计算机可读介质

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: 22744994

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: 22744994

Country of ref document: EP

Kind code of ref document: A1