Commodity identification method and device, medium and program product

Abstract

The invention relates to the field of computer data processing, in particular to a commodity recognition method, which comprises the following steps of: acquiring user generated content which comprises one or more of texts, pictures and videos; identifying characters in the text, the picture and the video user-generated content, and obtaining a commodity information named entity in the user-generated content, the named entity comprising commodity attributes; acquiring a context corresponding to the commodity information named entity; based on the commodity information context, performing multi-level recall on a predetermined commodity database at the same time, and selecting candidate items; and calculating the confidence of the candidate items based on the context, sorting the candidate items according to the confidence, and determining the option with the highest confidence as the commodity in the user generated content. The invention also relates to a commodity identification device, a medium and a program product. According to the method, multi-modal data is adopted, so that extraction of entity information is ensured; and a multi-level recall mode is adopted on the recall side, so that the recall efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the field of computer data processing, and in particular to a commodity identification method, device, medium and program product. Background Art

[0002] Users will post a lot of notes on the Internet, namely user-generated content (UGC), to express their opinions or interests. Most commonly, users will share their product experience with other users. In these notes, it will be mentioned what product the main content of the current note is about. The products mentioned in the notes can be linked to the smallest granular unit in our product library to increase the description of product information for the user who posted the note. For example, based on the products shared by users, we can favor the products when pushing product advertisements, link pushes, and promotional activities, thereby promoting transactions.

[0003] In the related technologies, the data source of recall is single and limited. In addition, the recall is mainly based on rules, and the recall efficiency is not high. Summary of the invention

[0004] The purpose of this application is to provide a commodity identification method, device, medium and program product, which adopts multimodal data in identification to ensure the extraction of entity information; and adopts a multi-level recall method on the recall side to improve the recall efficiency.

[0005] The embodiments of the present application disclose a method for identifying a product, the method comprising: obtaining user-generated content, the user-generated content comprising one or more of text, pictures and videos; identifying characters in the text, pictures and video user-generated content, obtaining product information named entities in the user-generated content, the named entities comprising product attributes; obtaining a context corresponding to the product information named entities; based on the product information context, simultaneously performing multi-level recall on a predetermined product database to select candidate items; and calculating the confidence of the candidate items based on the context, and sorting the candidate items according to the confidence, and determining the option with the highest confidence as the product in the user-generated content.

[0006] Optionally, obtaining the context corresponding to the named entity includes selecting one or more of a sentence, a title, and a topic in which the named entity is located to form the context.

[0007] Optionally, the multi-level recall includes a first-level recall and a second-level recall, wherein the priority of the first-level recall is higher than the priority of the second-level recall, and the predetermined rule of the first-level recall is that the word segmentation results in the product database are consistent with the named entity words in the entity relationship information, and the predetermined rule of the second-level recall is that the word segmentation results in the product database include the named entity words in the entity relationship information.

[0008] Optionally, the multi-level recall further includes a third-level recall, and a predetermined rule of the third-level recall is that more than 50% of the word segmentation results of the entity relationship information are included in the word segmentation results of the standardized product unit information in the commodity database.

[0009] Optionally, the commodity attributes include one or more of category, brand, and product series.

[0010] Optionally, calculating the confidence of the candidate item based on the context includes calculating the confidence of the candidate item based on a twin-tower model, the twin-tower model including a user-generated content side and a product side, the input of the user-generated content side includes the context, and the input of the product side is text in the product database.

[0011] Optionally, the input on the user-generated content side further includes pictures in the user-generated content, and the input on the product side further includes pictures in the product database.

[0012] Optionally, calculating the confidence of the candidate item based on the context includes calculating the confidence of the candidate item based on the context and discrete features of the candidate item.

[0013] An embodiment of the present application discloses an electronic device, characterized in that the device includes a processor and a memory storing computer-executable instructions, and the processor is configured to execute the instructions to implement the above-mentioned commodity identification method.

[0014] The embodiment of the present application discloses a computer-readable storage medium, characterized in that at least one computer instruction is stored in the computer-readable storage medium, and the at least one instruction is loaded and executed by a processor to implement the above-mentioned commodity identification method.

[0015] The embodiments of the present application disclose a computer program product, which includes computer instructions. When the computer instructions are executed, the above-mentioned commodity identification method is implemented.

[0016] The present application provides a commodity identification method, device, medium and program product, which obtains commodity information from text and pictures in a multimodal manner according to user-generated content, performs multi-level recall of a predetermined commodity library based on the commodity information, calculates the confidence of candidate items according to the context corresponding to the commodity information, and then sorts the candidate items according to the confidence, and finally determines the option with the highest confidence as the commodity in the user-generated content. The commodity identification method in the present application can use multimodal data to ensure the extraction of entity information; a multi-level recall method is used on the recall side to improve the recall efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of a scenario of a commodity identification method according to an embodiment of the present application;

[0018] Figure 2 is a flow chart of a commodity identification method according to an embodiment of the present application;

[0019] Figure 3 is a structural schematic diagram of a sorting method according to an embodiment of the present application;

[0020] Figure 4 It is a hardware structure block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] The present application will be further described below in conjunction with specific embodiments and accompanying drawings. It will be appreciated that the specific embodiments described herein are only intended to explain the present application and are not intended to limit the present application. In addition, for ease of description, only some but not all structures or processes related to the present application are shown in the accompanying drawings. It should be noted that in this specification, similar reference numerals and letters represent similar items in the following accompanying drawings.

[0022] It should be understood that although the terms "first", "second", etc. may be used herein to describe various features, these features should not be limited by these terms. These terms are used only to distinguish and should not be understood as indicating or implying relative importance. For example, a first feature may be referred to as a second feature, and similarly a second feature may be referred to as a first feature without departing from the scope of the exemplary embodiments.

[0023] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0024] Figure 1 It is a scene schematic diagram of a commodity identification method according to an embodiment of the present application.

[0025] like Figure 1As shown, the application scenario of the present application includes a client 100 and a server 200, and communication between the client 100 and the server 200 is achieved through a wired network or a wireless network connection.

[0026] It is worth noting that those skilled in the art will understand that there can be one or more clients 100, and the number of clients 100 should not be used as a limiting condition of the present application. The number of clients 100 is subject to the actual needs in the application scenario. For example, in a user-generated content recommendation system, the data source comes from different clients 100, and at least one client 100 sends data to the recommendation system. The client 100 can be a computer device such as a smart phone, a tablet computer, a user-generated content computer, a desktop computer, etc., which has computing power, input and output functions, and network connection functions. The server 200 can be a single server device, or it can be a computer device such as a server cluster, a cloud server, etc. that provides game application services to the outside.

[0027] In user-generated content, Dior lipstick may be mentioned in the text, but the specific series is not specified. Due to the lack of information, the single-modal approach cannot complete the link to the product library SPU. However, it can be read from the OCR results and image information that the product series is the 999 series. Therefore, in related technologies, the data source for recall is relatively single, and most of them only consider the text, not the OCR information in the picture, which is quite limited.

[0028] In addition, the recall in the prior art is mainly based on rules, and the recall efficiency is not high.

[0029] In response to the above problems, the present application provides a product identification method, device, medium and program product, which can use multimodal data to ensure the extraction of entity information; and use a multi-level recall method on the recall side to improve recall efficiency.

[0030] Figure 2 It is a flow chart of a commodity identification method according to an embodiment of the present application.

[0031] like Figure 2 As shown, the commodity identification method provided in one embodiment of the present application requires Figure 1 The method is implemented by the cooperation of the client 100 and the server 200, and includes:

[0032] In S100 , user-generated content is obtained, where the user-generated content includes one or more of text, picture, and video.

[0033] Obtain user generated content, which may be text and pictures. In some embodiments, the user generated content may also include videos.

[0034] In S200 , text in user-generated content is recognized, and named entities in the user-generated content are acquired, where the named entities include product attributes.

[0035] Users can express their experience of using a product or their preference for a product in user-generated content, so user-generated content is associated with one or more products. For example, if user A mentions a certain brand of mobile phone in the text part of a note and discloses the specific model of the mobile phone in the text tag in the picture part of the note, then this note contains product information of the brand and model of mobile phone.

[0036] The named entities including product attributes in the user-generated content can be determined based on the text in the user-generated content, such as text and text in pictures. The text in the picture can be obtained through OCR technology. For example, in the picture, the image data can be divided first, the text information therein can be identified through OCR, or the corresponding text information can be obtained based on the feature vector of the image itself. In some examples, when the user-generated content includes a video, the video frame cutting method can be used to extract text information from each frame or certain specific frames of the picture.

[0037] In some embodiments, entity recognition tools, such as Stanford CoreNLP or pyltp, can be used to identify named entities in user-generated content. Stanford CoreNLP is a natural language processing tool, and its main functions include word segmentation, part-of-speech tagging, named entity recognition, syntactic structure analysis, and dependency analysis. As a natural language processing tool, pyltp is a Python encapsulation of LTP, which provides functions of word segmentation, part-of-speech tagging, named entity recognition, dependency syntactic analysis, and semantic role labeling.

[0038] Named entity recognition refers to the identification of entities with specific meanings in text, mainly including names of people, places, proper nouns, organization names, etc. Generally speaking, the task of named entity recognition is to identify three major categories of named entities in the text to be processed, such as entity category, time category and number category, as well as seven subcategories, such as names of people, organization names, place names, time, date, currency and percentage. Named entity recognition aims to identify entities in a string of text and mark the type it refers to, such as names of people, places, etc. Specifically, given an input sentence, named entity recognition requires entity annotation for each token in the sentence, such as names of people, organizations / institutions, place names, dates, etc. The token here is the smallest semantic unit.

[0039] The recognition result of the named entity is used as a candidate word, and special symbols are removed, and uppercase English and traditional Chinese characters are converted and other preprocessing is performed. In an example, a user publishes a content, and the user-generated content includes text and pictures, where the text content mainly introduces women's lipstick, and the picture content is a lipstick photo and a text watermark. The server 200 makes entity annotations for each token in the user-generated content, identifies the named entity in the user-generated content, and can obtain the product "lipstick" from the text, and can obtain the brand and series corresponding to the lipstick according to the picture content.

[0040] In some implementations, a model may be trained for entity recognition, and the recognition result may be used as a candidate word. In entity recognition, special symbols may be removed, and uppercase English and traditional Chinese characters may be converted, and other preprocessing may be performed.

[0041] In some embodiments, the commodity attributes of the named entities in the user-generated content include one or more of categories, brands, and product series. Among them, the category can be a specific type of a certain commodity, for example, the categories under electronic products include mobile phones or laptops, and the categories under cosmetics include lipstick or liquid foundation. In addition, the brand can be a specific brand to which a certain commodity belongs, and a series of a certain category of products of a certain brand can determine a specific commodity. In this way, the named entity including the commodity attributes can determine the product unit of the commodity. For example, a certain commodity is mentioned in the user's notes, and the text or the text of the picture includes the named entities "aa brand" and "bb category", then this group of named entities determines the product unit of the "aa brand-bb category" commodity.

[0042] The Standardized Product Unit (SPU) is the smallest unit of commodity information aggregation, a set of reusable, easily retrievable standardized information that describes the characteristics of a product. In other words, the same SPU is called a commodity, which refers to the smallest aggregation unit of a commodity. For example, in the category structure of commodities, the data of a certain model of mobile phone belongs to the category of mobile phone → smart phone → a certain brand of mobile phone → a certain model, then the category information can be used to represent the SPU. In some examples, the SPU of the mobile phone can be: "aa brand bb mobile phone cc series". A commodity category database can be established based on the SPU.

[0043] In some embodiments, the named entities in the user-generated content may correspond to the SPU. For example, if a certain product is mentioned in the user's notes, and the text or the text of the picture includes the named entities "aa brand", "bb category", and "cc series", then this set of named entities determines the SPU of the product "aa brand-bb category-cc series". In some other examples, if a certain cosmetic is mentioned in the user's notes, and the text or the text of the picture includes the named entities "dd brand", "ee category", "ff series", "gg color", and "hh texture", then this set of named entities determines the SPU corresponding to this cosmetic.

[0044] In some implementations, recognizing text in user-generated content and obtaining named entities in the user-generated content further includes the following steps.

[0045] In S201 , a plurality of named entities are identified based on text in user-generated content.

[0046] The text in user-generated content can include multiple named entities. For example, a user's note M mentions a certain product, and also mentions the brand and series of the product. For another example, a user's note N mentions more than one product, such as product A and product B. The note also discloses the AA brand, BB brand, AAA series, BBB series, CC time, DD place name, etc.

[0047] Based on the entity recognition tool described above, all named entities including product attributes in note N can be identified, namely AA brand, BB brand, AAA series, and BBB series.

[0048] In S202, a plurality of named entities are classified.

[0049] Based on multiple named entities in user-generated content, a relation extraction model can be used to determine the relationship between multiple named entities, that is, to classify multiple named entities. For example, the AA brand and the BB brand in note N are one category, and the AAA series and the BBB series are another category.

[0050] The named entities in S201 and the classification in S202 can be implemented by separate models. For example, the named entities are extracted based on the user's notes by a sequence labeling model, and the relationship between the named entity pairs is obtained by a relational classification model. In some other examples, the same model can be used to complete S201 and S202.

[0051] In S203, entity relationship information is obtained based on the classification, wherein each entity relationship information corresponds to a commodity unit.

[0052] Entity relationship information (mention) can be obtained by classifying multiple named entities, and each mention corresponds to a commodity unit. For example, in note N, A commodity-AA brand-AAA series is one mention, and B commodity-BB brand-BBB series is another mention.

[0053] In S300, a context corresponding to the named entity is obtained.

[0054] In some implementations, the context corresponding to the named entity includes one or more of the sentence, title, and topic in which the named entity is located.

[0055] Acquiring the context corresponding to the named entity is to extract the content related to the named entity description with commodity attributes. In some embodiments, the sentence where the named entity word is located, the title, topic, etc. can be selected and combined into the context according to the identified named entity word.

[0056] In S400 , based on the context, a multi-level recall is performed on the predetermined product database simultaneously to select candidate items.

[0057] In some embodiments, the multi-level recall includes a first-level recall and a second-level recall, wherein the priority of the first-level recall is higher than that of the second-level recall. The predetermined rule for the first-level recall is that the word segmentation results in the commodity database and the named entity words in the entity relationship information are consistent. The predetermined rule for the second-level recall is that the word segmentation results include the named entity words in the entity relationship information. In other words, the first-level recall is an accurate recall based on a predetermined rule, and the second-level recall is a full-token matching recall.

[0058] In some examples, the first-level recall is an accurate recall. For example, the entity relationship information is exactly the same as the SPU name or alias in the product library. For example, the entity relationship information is: Dior-999-lipstick, and the SPU name in the product library is: dior999 lipstick, which can be accurately recalled.

[0059] In some examples, the second-level recall is a full token match recall, that is, the word segmentation results of the entity relationship information are included in the word segmentation results of the SPU name in the product library. For example, if the entity relationship information is: L'Oreal-Foundation Liquid, and the SPU name in the product library is: L'Oreal Concealer Whitening Foundation Liquid, it can be recalled by full token matching.

[0060] In some embodiments, the multi-level recall also includes a third-level recall, which is similar to a synonym recall or a hyponym recall. The predetermined rule for the third-level recall is that more than 50% of the segmentation results of the entity relationship information are included in the segmentation results of the SPU in the product database. For example, the entity relationship information is Estee Lauder-Little Brown Bottle-Eye Cream, and the name of the SPU in the product database is: Estee Lauder Little Brown Bottle Anti-wrinkle Whitening, which can be fuzzy recalled.

[0061] The use of multi-level recall on the recall side can avoid the problem of non-standard product SPU names. Token recall and fuzzy recall can increase the number of recall candidates, avoid missed recalls, and increase the robustness of the recall.

[0062] In some embodiments, the combination logic of the multi-level recall is executed sequentially, and finally a predetermined number of candidate items are selected. For example, the result of the recall is first-level recall first, then second-level recall, and finally third-level recall.

[0063] In one example, the system needs to recall 20 candidate items, and the first-level recall, second-level recall and third-level recall are executed sequentially. The result of the first-level recall is to recall 10 candidate items, and the recall result does not meet the quantity requirement. Then the second-level recall is performed, and the result of the second-level recall is to recall 5 candidate items, and the recall result still does not meet the quantity requirement. The third-level recall needs to be continued, and the result of the third-level recall is to recall 5 candidate items, and the multi-level recall task is completed.

[0064] In another example, the system needs to recall 20 candidate items. The first-level recall, second-level recall and third-level recall are executed sequentially. The result of the first-level recall is to recall 10 candidate items, and the recall result does not meet the quantity requirement. Then the second-level recall is performed, and the result of the second-level recall is to recall 10 candidate items. The recall result meets the quantity requirement. There is no need to continue the third-level recall, and the multi-level recall task is completed.

[0065] In another example, the system needs to recall 20 candidate items. The first-level recall, second-level recall and third-level recall are executed sequentially. The result of the first-level recall is to recall 20 candidate items. The recall result meets the recall quantity requirement. There is no need to continue to execute the second-level recall and the third-level recall, and the multi-level recall task is completed.

[0066] In S500 , the confidence of the candidate items is calculated based on the context, and the candidate items are sorted according to the confidence, and the option with the highest confidence is determined as the standardized product unit in the user-generated content.

[0067] Combined with the context of the named entity, all candidate products are ranked according to the calculated confidence.

[0068] In some embodiments, the confidence is calculated using the structure of a twin tower model. The twin towers in the twin tower model refer to the user-generated content side and the product side. The twin tower model as a whole includes an input layer, a representation layer, and a matching layer. In the input layer, the input of the user-generated content side includes the context extracted from the user-generated content, and the input of the product side is the text information of the product in the product database. The context here includes the context corresponding to the named entities in the text of the user-generated content, and may also include the context of the named entities in the text obtained in the picture through OCR (optical character recognition) technology.

[0069] In some other embodiments, the input layer may also include pictures. For example, the input on the user-generated content side may also include pictures in the user-generated content and / or screenshots in the video. Pictures are images contained in the user-generated content itself. If the user-generated content is in the form of a video, a key frame may be cut out every five seconds in the video and the key frame may be used as picture information. The input on the product side may include multiple pictures under the product category.

[0070] The representation layer in the dual-tower model can convert the input into a vector, and the matching layer can calculate the similarity of the vectors on the user-generated content side and the product side.

[0071] The commodity data in the dual-tower model can be calculated offline without affecting the online real-time performance.

[0072] In some embodiments, calculating the confidence of the candidate based on the context includes calculating the confidence of the candidate based on the context and discrete features of the candidate. Discrete features can be added to the matching layer of the twin tower model to increase the robustness of the model. Figure 3 It is a structural diagram of a sorting method according to an embodiment of the present application.

[0073] like Figure 3 As shown in FIG. 1 , a dual-tower model is used to perform the sorting operation. The feature codes input from the user-generated content side (left side) and the product side (right side) are concatenated into vectors, and then the similarity is obtained by calculating the cosine distance between the two vectors.

[0074] On the left side of the twin tower model, two BERT models are used to encode the context of named entities in user-generated content and OCR text in pictures in user-generated content, respectively. In addition, a ResNet model is used to encode the features of pictures in user-generated content. The vectors obtained from the three paths are concatenated, and the self-attention mechanism is used to consider the correlation between different vectors. When the user-generated content includes videos, key frames can be cut out of the video at each time interval, such as 5 seconds, and the key frames can be used as input pictures.

[0075] On the right side of the twin tower model, the BERT model is used to feature encode the text in the SPU information, and the ResNet model is used to feature encode multiple product images under this SPU. The obtained vectors are cascaded, and the self-attention mechanism is used to consider the correlation between different vectors. The number of product images can be set by yourself, such as 10, 20 or 30.

[0076] As an implementation method, the confidence of the candidate items can be calculated based on discrete features. Discrete features are used as intermediate inputs for confidence scoring. For example, the degree of literal overlap can be calculated by measuring how much of the text information of the SPU appears in the context information. Or the popularity of the current SPU can be increased, that is, the inventory, sales volume, and merchant information of the product under the current SPU. The added discrete features can ensure the ability of manual intervention and increase the robustness of the link.

[0077] exist Figure 3 In the dual-tower model, user-generated content information and product information are input, plus a little discrete feature, and finally the label label and confidence score score are output. The label indicates whether the current user-generated content is linked to the product 0 or 1, and the confidence indicates the probability of connecting to a certain SPU in the product library. For example, based on the input on the left and right sides, the label between the user-generated content and SPU1 in the product library is 1, and the confidence is 0.9, while the similarity between the user-generated content and SPU2 in the product library meets a certain threshold, then the label is 1, but the confidence is 0.3. In this case, the user-generated content is linked to SPU1. If the label between the user-generated content and SPU1 in the product library is 0 according to the dual-tower model, select another SPU in the candidate and repeat the above input, comparison, label and confidence steps according to the dual-tower model.

[0078] Figure 4 It is a hardware structure block diagram of an electronic device implementing an embodiment of the present application.

[0079] like Figure 4 As shown, the electronic device 400 may include one or more processors 402, a system motherboard 408 connected to at least one of the processors 402, a system memory 405 connected to the system motherboard 408, a non-volatile memory (NVM) 406 connected to the system motherboard 408, and a network interface 410 connected to the system motherboard 408.

[0080] The processor 402 may include one or more single-core or multi-core processors. The processor 402 may include any combination of a general-purpose processor and a dedicated processor (e.g., a graphics processor, an application processor, a baseband processor, etc.). In an embodiment of the present application, the processor 402 may be configured to execute the following: Figure 2 The product identification method shown.

[0081] In some embodiments, system board 408 may include any suitable interface controller to provide any suitable interface to at least one of processors 402 and / or any suitable device or component in communication with system board 408 .

[0082] In some embodiments, the system board 408 may include one or more memory controllers to provide an interface to the system memory 405. The system memory 405 may be used to load and store data and / or instructions. In some embodiments, the system memory 405 of the electronic device 400 may include any suitable volatile memory, such as a suitable dynamic random access memory (DRAM).

[0083] NVM 406 may include one or more tangible, non-transitory computer-readable media for storing data and / or instructions. In some embodiments, NVM 406 may include any suitable non-volatile memory such as flash memory and / or any suitable non-volatile storage device, such as at least one of a HDD (Hard Disk Drive), a CD (CompactDisc) drive, and a DVD (Digital Versatile Disc) drive.

[0084] NVM 406 may include a portion of storage resources installed on a device of electronic device 400, or it may be accessible by the device but not necessarily a portion of the device. For example, NVM 406 may be accessed over a network via network interface 410.

[0085] In particular, system memory 405 and NVM 406 may include a temporary copy and a permanent copy of instruction 420, respectively. Instruction 420 may include instructions that, when executed by at least one of processors 402, cause electronic device 400 to implement the following: Figure 2 In some embodiments, instructions 420, hardware, firmware and / or software components thereof may additionally / alternatively be located in system board 408, network interface 410 and / or processor 402.

[0086] The network interface 410 may include a transceiver for providing a radio interface for the electronic device 400, thereby communicating with any other suitable device (e.g., a front-end module, an antenna, etc.) through one or more networks. In some embodiments, the network interface 410 may be integrated with other components of the electronic device 400. For example, the network interface 410 may be integrated with at least one of the processor 402, the system memory 405, the NVM 406, and a firmware device (not shown) having instructions. When at least one of the processors 402 executes the instructions, the electronic device 400 implements Figure 2 One or more of the various embodiments shown.

[0087] The network interface 410 may further include any suitable hardware and / or firmware to provide a multiple-input multiple-output radio interface. For example, the network interface 410 may be a network adapter, a wireless network adapter, a telephone modem, and / or a wireless modem.

[0088] In one embodiment, at least one of the processors 402 may be packaged together with one or more controllers for the system board 408 to form a system in package (SiP). In one embodiment, at least one of the processors 402 may be integrated on the same die with one or more controllers for the system board 408 to form a system on a chip (SoC).

[0089] The electronic device 400 may further include an input / output (I / O) device 412 connected to the system motherboard 408. The I / O device 412 may include a user interface to enable a user to interact with the electronic device 400; the design of the peripheral component interface enables the peripheral components to interact with the electronic device 400. In some embodiments, the electronic device 400 further includes a sensor for determining at least one of an environmental condition and location information related to the electronic device 400.

[0090] In some embodiments, I / O device 412 may include, but is not limited to, a display (e.g., an LCD display, a touch screen display, etc.), a speaker, a microphone, one or more cameras (e.g., a still image camera and / or a video camera), a flashlight (e.g., an LED flash), and a keyboard.

[0091] In some embodiments, the peripheral component interface may include, but is not limited to, a non-volatile memory port, an audio jack, and a power interface.

[0092] It is to be understood that the structure illustrated in the embodiment of the present application does not constitute a specific limitation on the electronic device 400. In other embodiments of the present application, the electronic device 400 may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0093] Program code can be applied to input instructions to perform the functions described in this application and generate output information. The output information can be applied to one or more output devices in a known manner. For the purposes of this application, a system for processing instructions including processor 402 includes any system having a processor such as a digital signal processor (DSP), a microcontroller, an application specific integrated circuit (ASIC), or a microprocessor.

[0094] Program code can be implemented with high-level programming language or object-oriented programming language to communicate with the processing system. When necessary, program code can also be implemented with assembly language or machine language. In fact, the mechanism described in this application is not limited to the scope of any specific programming language. In either case, the language can be a compiled language or an interpreted language.

[0095] According to an embodiment of the present application, a computer-readable storage medium is further proposed, wherein at least one computer instruction is stored in the computer-readable storage medium, and the at least one instruction is loaded and executed by a processor to implement the aforementioned commodity identification method.

[0096] According to an embodiment of the present application, a computer program product is also proposed. The computer program product includes computer instructions. When the computer instructions are executed, the aforementioned commodity identification method is implemented.

[0097] The illustrative embodiments of the present application include, but are not limited to, a commodity identification method, apparatus, medium, and program product.

[0098] The various aspects of the illustrative embodiments will be described using terms commonly used by those skilled in the art to convey the essence of their work to other persons skilled in the art. However, it is apparent to those skilled in the art that some alternative embodiments may be implemented using the features described in part. For purposes of explanation, specific numbers and configurations are set forth to provide a more thorough understanding of the illustrative embodiments. However, it is apparent to those skilled in the art that alternative embodiments may be implemented without the specific details. In some other cases, some well-known features are omitted or simplified herein to avoid obscuring the illustrative embodiments of the present application.

[0099] In addition, various operations will be described as multiple separate operations in a manner that is most helpful for understanding the illustrative embodiments; however, the order of description should not be interpreted as implying that these operations must rely on the order of description, and many of the operations can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can also be rearranged. When the described operations are completed, the process can be terminated, but can also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0100] References in the specification to "an example", "an example", "an embodiment", "an implementation", etc. indicate that the described embodiment may include a particular feature, structure, or property, but each embodiment may or may not necessarily include the particular feature, structure, or property. Moreover, these phrases are not necessarily directed to the same embodiment. In addition, when particular features are described in conjunction with a specific embodiment, the knowledge of those skilled in the art can affect the combination of these features with other embodiments, whether or not these embodiments are explicitly described.

[0101] Unless the context dictates otherwise, the terms "comprising," "having," and "including" are synonymous. The phrase "A and / or B" means "(A), (B), or (A and B)."

[0102] As used herein, the term "module" may refer to, be part of, or include: memory (shared, dedicated, or group) for running one or more software or firmware programs, application-specific integrated circuits (ASICs), electronic circuits and / or processors (shared, dedicated, or group), combinational logic circuits, and / or other suitable components that provide the described functionality.

[0103] In the accompanying drawings, some structural or method features may be shown in a specific arrangement and / or order. However, it should be understood that such a specific arrangement and / or order is not necessary. Instead, in some embodiments, these features may be described in a manner and / or order different from that shown in the illustrative drawings. In addition, the structural or method features included in a specific figure do not mean that all embodiments need to include such features. In some embodiments, these features may not be included or may be combined with other features.

[0104] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0105] Similarly, it should be understood that in order to streamline the present application and aid in understanding one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present application, the various features of the present application are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the following intention: the claimed application requires more features than the features explicitly recited in each claim. More specifically, as reflected in the claims, the inventive aspects lie in less than all the features of the single embodiment disclosed above. Therefore, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, with each claim itself serving as a separate embodiment of the present application.

[0106] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0107] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present application and form different embodiments. For example, in the claims, any one of the claimed embodiments can be used in any combination.

Claims

1. A commodity identification method, used in electronic equipment, characterized in that: include: Obtaining user-generated content, where the user-generated content includes one or more of text, pictures, and videos; Recognize text in the user-generated content and obtain named entities in the user-generated content, wherein the named entities include product attributes; Obtaining a context corresponding to the named entity; Based on the context, a predetermined product database is simultaneously recalled in multiple stages to select candidate items; as well as The confidences of the candidate items are calculated based on the context, and the candidate items are sorted according to the confidences, and the option with the highest confidence is determined as the product in the user-generated content.

2. The method according to claim 1, characterized in that Recognizing the text in the user-generated content and obtaining the named entities in the user-generated content further includes: identifying a plurality of named entities based on the text; classifying the plurality of named entities, and Entity relationship information is acquired based on the classification, wherein each entity relationship information corresponds to a commodity unit.

3. The method according to claim 1, characterized in that Acquiring the context corresponding to the named entity includes selecting one or more of the sentence, title, and topic where the named entity is located to form the context.

4. The method according to claim 2, characterized in that: The multi-level recall includes a first-level recall and a second-level recall, wherein the priority of the first-level recall is higher than that of the second-level recall, and the predetermined rule of the first-level recall is that the word segmentation results in the commodity database are consistent with the named entity words in the entity relationship information, and the predetermined rule of the second-level recall is that the word segmentation results in the commodity database include the named entity words in the entity relationship information.

5. The method according to claim 4, characterized in that The multi-level recall also includes a third-level recall, and the predetermined rule of the third-level recall is that more than 50% of the word segmentation results of the entity relationship information are included in the word segmentation results of the standardized product unit information in the commodity database.

6. The method according to claim 1, characterized in that The commodity attributes include one or more of category, brand, and product series.

7. The method according to claim 1, characterized in that Calculating the confidence of the candidate item based on the context includes calculating the confidence of the candidate item based on a twin-tower model, the twin-tower model includes a user-generated content side and a product side, the input of the user-generated content side includes the context, and the input of the product side is text in the product database.

8. The method according to claim 7, characterized in that The input on the user-generated content side also includes pictures in the user-generated content, and the input on the product side also includes pictures in the product database.

9. The method according to claim 7, characterized in that: Calculating the confidence of the candidate item based on the context includes calculating the confidence of the candidate item based on the context and discrete features of the candidate item.

10. An electronic device, characterized in that: The device comprises a memory storing computer executable instructions and a processor, wherein the processor is configured to execute the instructions to implement the method according to any one of claims 1-9.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one computer instruction, and the at least one instruction is loaded and executed by a processor to implement the method according to any one of claims 1 to 9.

12. A computer program product, characterized in that The computer program product comprises computer instructions, and when the computer instructions are executed, the method according to any one of claims 1 to 9 is implemented.

Images