CN110222167B

CN110222167B - Method and system for acquiring target standard information

Info

Publication number: CN110222167B
Application number: CN201910595218.9A
Authority: CN
Inventors: 马良庄
Original assignee: Advanced New Technologies Co Ltd
Current assignee: Advanced New Technologies Co Ltd; Advantageous New Technologies Co Ltd
Priority date: 2019-07-03
Filing date: 2019-07-03
Publication date: 2023-04-07
Anticipated expiration: 2039-07-03
Also published as: CN110222167A

Abstract

The embodiment of the specification relates to a method and a system for acquiring target standard information, and belongs to the technical field of artificial intelligence. The method comprises the following steps: obtaining a question of a target user; determining a first set of standard information based on the target user question and a first preset algorithm; acquiring a track factor of a target user; determining first type of information related to the standard information based on the target user track factor and a second preset algorithm; one or more target criteria information are determined based on the first set of criteria information and the first type of information associated with the criteria information.

Description

Method and system for acquiring target standard information

Technical Field

One or more embodiments of the present disclosure relate to the field of artificial intelligence technology, and more particularly, to a method and system for automatically determining target standard information.

Background

In intelligent customer service, a robot automatically processes information input by a client to determine the matters the client wants to consult or the answers of related questions that the client intends to know. The customer usually uses the text information input by the user to perform matching identification in the process of interacting with the robot. The text information typically includes textual question statements in a single turn of a dialog by the user, as well as contextual information in multiple turns of the dialog. In most cases, during the interaction process between the user and the robot, the user may use simplified dialogue for interaction, which results in that the text information input by the user may be fuzzy text information, and it is difficult to accurately match the text information to the accurate information when performing matching recognition. For example, the user may enter "audit," and it is difficult for the interactive system to match to identify what accounts the user specifically needs to audit, or the scope of the audit, etc. There is therefore a need to provide a more accurate question and answer engine.

Disclosure of Invention

One of one or more embodiments in this specification provides a method of obtaining target criteria information, the method comprising: obtaining a question of a target user; determining a first set of standard information based on the question and a first preset algorithm; obtaining a track factor of the target user, wherein the track factor reflects at least one behavior of the target user on one or more service platforms; determining first type of information related to standard information based on the track factor and a second preset algorithm; one or more target criteria information are determined based on the first set of criteria information and the first type of criteria information-related information.

In some embodiments, the method of obtaining target criteria information further comprises outputting one or more target criteria information.

In some embodiments, the method of obtaining target criteria information further comprises: obtaining feedback of a user on the one or more target standard information; and updating the first preset algorithm and/or the second preset algorithm according to the feedback of the user.

In some embodiments, the first preset algorithm and/or the second preset algorithm comprises a machine learning model, the machine learning model comprising a Bert model.

In some embodiments, determining one or more target criteria information based on the first set of criteria information and the first type of criteria information-related information comprises: screening a first group of standard information based on the first type of information related to the standard information to determine one or more target standard information; the first type of information related to the labeling information comprises screening auxiliary information, and the screening auxiliary information reflects screening conditions of the target standard information.

In some embodiments, the first type of information related to standard information comprises a second set of standard information; the second set of standard information includes one or more standard information and/or content information corresponding thereto.

In some embodiments, said determining one or more target criteria information based on said first set of criteria information and said first type of annotation information-related information comprises: determining common standard information of the first group of standard information and the second group of standard information; one or more target criteria information is determined from the common criteria information.

In some embodiments, the first set of criteria information further includes a first numerical value corresponding to each of its criteria information; the second set of criteria information further includes a second value corresponding to each of the criteria information.

In some embodiments, determining one or more target criteria information based on the first set of criteria information and the first type of annotation information-related information comprises: determining common standard information of the first group of standard information and the second group of standard information; operating the first numerical value and the second numerical value of the common standard information; and determining one or more target standard information from the common standard information based on the operation result.

In some embodiments, the operation comprises a multiplication or an addition.

One of the embodiments of the present specification provides a system for acquiring target standard information, the system including: the first acquisition module is used for acquiring the question of the target user; the first determining module is used for determining a first group of standard information based on the problem and a first preset algorithm; the second acquisition module is used for acquiring the track factor of the target user; the trajectory factor reflects at least one behavior of the target user on one or more service platforms; the second determining module is used for determining the first type of information related to the standard information based on the track factor and a second preset algorithm; a target determination module that determines one or more target criteria information based on the first set of criteria information and the first type of criteria information-related information.

In some embodiments, the system further comprises: and the target output module is used for outputting one or more pieces of target standard information.

In some embodiments, the system further comprises: the third acquisition module is used for acquiring feedback of the user on the one or more target standard information; and the algorithm optimization module is used for updating the first preset algorithm and/or the second preset algorithm according to the feedback of the user.

In some embodiments, the target determination module is further configured to filter a first set of criteria information based on the first type of information related to criteria information, and determine one or more target criteria information; the first type of information related to the standard information includes screening auxiliary information reflecting a screening condition of the target standard information.

In some embodiments, the target determination module is further configured to determine common criteria information of the first set of criteria information and the second set of criteria information; one or more target criteria information is determined from the common criteria information.

In some embodiments, the goal determination module is further to: determining common standard information of the first group of standard information and the second group of standard information; operating the first numerical value and the second numerical value of the common standard information; and determining one or more target standard information from the common standard information based on the operation result.

In some embodiments, the operation comprises a multiplication or an addition.

One of the embodiments of the present specification provides an apparatus for obtaining target criteria information, comprising a processor and a memory, the memory for storing computer instructions; the processor is configured to execute at least a portion of the computer instructions to implement the operations described in any of the embodiments in this specification.

Drawings

One or more embodiments of the present specification are further illustrated by way of example embodiments, which are described in detail below and illustrated in the accompanying drawings. These embodiments are not intended to be limiting, and in these embodiments like numerals are used to indicate like structures, wherein:

FIG. 1 is a schematic diagram of an application scenario for obtaining target criteria information, according to some embodiments of the present description;

FIG. 2 is a block diagram of an exemplary system shown in accordance with some embodiments of the present description;

FIG. 3 is an exemplary flow diagram illustrating a method for obtaining target criteria information according to some embodiments of the present description;

FIG. 4 is a schematic diagram of a sub-flow diagram illustrating determining target criteria information according to further embodiments of the present description;

FIG. 5 is a diagrammatic illustration of a training flow of a machine learning model in accordance with a first pre-set algorithm shown in some embodiments of the present description;

FIG. 6 is a diagrammatic illustration of a training flow of a machine learning model in accordance with a second pre-set algorithm shown in some embodiments of the present description;

FIG. 7 is an illustration of a sub-flow shown to obtain a trajectory factor of a target user in accordance with some embodiments of the present description.

Detailed Description

In order to more clearly illustrate the technical solutions of one or more embodiments of the present disclosure, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only examples or embodiments of the present description, and that for a person skilled in the art, one or more embodiments of the present description can also be applied to other similar scenarios according to these drawings without inventive effort. Unless otherwise apparent from the context, or otherwise indicated, like reference numbers in the figures refer to the same structure or operation.

It should be understood that "system," "device," "unit," and/or "module" as used herein is a method for distinguishing between different components, elements, parts, portions, or assemblies of different levels. However, other words may be substituted by other expressions if they accomplish the same purpose.

As used in one or more embodiments of the present specification and in the claims, the terms "a," "an," and/or "the" are not intended to be inclusive in the singular, but rather are intended to be inclusive in the plural, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that steps and elements are included which are explicitly identified, that the steps and elements do not form an exclusive list, and that a method or apparatus may include other steps or elements.

Flow diagrams are used in one or more embodiments of the present description to illustrate the operations performed by systems according to one or more embodiments of the present description. It should be understood that the preceding or following operations are not necessarily performed in the exact order in which they are performed. Rather, the various steps may be processed in reverse order or simultaneously. Meanwhile, other operations may be added to the processes, or a certain step or several steps of operations may be removed from the processes.

One or more embodiments of the present description may be applied to various customer service question and answer systems or search engine systems, etc. The different customer service question-answering systems include but are not limited to one or a combination of financial, shopping, traveling, education, medical treatment and the like. For example, a customer service question and answer system using a machine question and answer such as an online shopping service, a bank service, a payment platform service, a shopping guide service in a shopping mall, a ticket ordering service, a convenience service, an education consultation service, and a guide service. The search engine system includes, but is not limited to, one or a combination of several of a financial platform search engine, a shopping platform search engine, a travel platform search engine, an education platform search engine, a medical platform search engine, a knowledge sharing platform search engine, and the like.

Different embodiment application scenarios of one or more embodiments of the present specification include, but are not limited to, one or a combination of several of a web page, a browser plug-in, a client App, a customization system, an intra-enterprise analysis system, an artificial intelligence robot, and the like. It should be understood that the application scenarios of the system and method of one or more embodiments of the present specification are only examples of one or more embodiments of the present specification, and it will be apparent to those of ordinary skill in the art that one or more embodiments of the present specification can also be applied to other similar scenarios according to these drawings without inventive effort. For example, other similar help guidance systems.

Fig. 1 is a schematic diagram of an application scenario of a target standard information acquisition and presentation system 100 according to one or more embodiments of the present description. The target standard information acquisition and presentation system 100 may be used to provide information such as bank customer service questions and answers, payment platform customer service questions and answers, mall shopping guide customer service questions and answers, ticketing customer service questions and answers, convenience service customer service questions and answers, education consultation customer service questions and answers, referral customer service questions and answers, and the like. The target standard information acquisition and presentation system 100 may include a server 110, a storage device 120, a user terminal 130, and a network 140.

Server 110 may be configured to process information and/or data related to a query entered by a user. For example, the server 110 may transform the query into a query vector. As another example, the server 110 may determine a set of criteria information from a criteria information base based on a predetermined algorithm. The server 110 may also be configured to process information and/or data related to the user's behavior and/or personal attribute information. For another example, the server 110 may obtain a trajectory factor in response to the user's behavior and/or personal attribute information, and determine a set of criterion information based on the trajectory factor and a preset algorithm. In some embodiments, the server 110 may be a single server or a group of servers. The set of servers can be centralized or distributed (e.g., the servers 110 can be a distributed system). In some embodiments, the server 110 may be local or remote. For example, server 110 may access information and/or data stored on user terminal 130 or storage device 120 via network 140. As another example, server 110 may be coupled to user terminal 130 and/or storage device 120 to access stored information and/or data. In some embodiments, the server 110 may be implemented on a cloud platform. By way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-tiered cloud, and the like, or any combination thereof.

Storage device 120 may store data and/or instructions. For example, the storage device 120 may store a pre-generated machine learning model. As another example, the storage device 120 may store a criteria information base and/or at least two historical user input-criteria information pairs and/or at least two historical track factor-criteria information pairs. As another example, the storage device 120 may store historical behavior of one or more users at the service platform and/or personal attributes of the users. As another example, storage device 120 may store data and/or instructions that server 110 may perform or be used to perform the example methods described in one or more embodiments of this specification. In some embodiments, storage device 120 may include mass storage, removable storage, volatile read-write memory, read-only memory (ROM), and the like, or any combination thereof. Exemplary mass storage may include magnetic disks, optical disks, solid state disks, and the like. Exemplary removable memories may include flash drives, floppy disks, optical disks, memory cards, compact disks, magnetic tape, and so forth. Exemplary volatile read-write memories can include Random Access Memory (RAM). Exemplary RAM may include Dynamic Random Access Memory (DRAM), double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), static Random Access Memory (SRAM), thyristor random access memory (T-RAM), zero-capacitance random access memory (Z-RAM), and the like. Exemplary read-only memories may include mask read-only memory (MROM), programmable read-only memory (PROM), erasable programmable read-only memory (perrom), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM), digital versatile disc read-only memory, and the like. In some embodiments, the storage device 120 may be implemented on a cloud platform. By way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-tiered cloud, and the like, or any combination thereof.

The user terminal 130 may comprise an application side of the target criteria information acquisition and presentation system 100. The user terminal 130 may also be used to receive query data of a client or user, and output data corresponding to the query data. In some embodiments, the user's query data may include the user's questions or questions. The user's questioning means may include various forms such as one or a combination of text input, voice input, image information input, and the like. In some embodiments, the output data corresponding to the query data may include standard questions corresponding to the query data and/or content information corresponding to the standard questions. In some embodiments, the content information corresponding to the standard question may be viewed as an answer to the question. User terminal 130 may include any electronic device used by a user. In some embodiments, the user terminal 130 may be a mobile device 130-1, a tablet computer 130-2, a laptop computer 130-3, a desktop computer 130-4, the like, or any combination thereof. In some embodiments, the mobile device 130-1 may include a wearable apparatus, a smart mobile device, a virtual reality device, an augmented reality device, and the like, or any combination thereof. In some embodiments, the wearable device may include a smart bracelet, a smart footwear, smart glasses, a smart helmet, a smart watch, a smart garment, a smart backpack, a smart accessory, or the like, or any combination thereof. In some embodiments, the smart mobile device may include a smart phone, a Personal Digital Assistant (PDA), a gaming device, a navigation device, a point of sale (POS), or the like, or any combination thereof. In some embodiments, the virtual reality device and/or the enhanced virtual reality device may include a virtual reality helmet, virtual reality glasses, virtual reality eyecups, augmented reality helmets, augmented reality glasses, augmented reality eyecups, and the like, or any combination thereof. For example, the virtual reality device and/or augmented reality device may include Googleglass, riftCon, fragmentsTM, gearVRTM, and the like. In some embodiments, desktop computer 130-4 may be an on-board computer, an on-board television, or the like.

In some embodiments, user terminal 130 may include at least one network port. The at least one network port may be configured to send information to and/or receive information from one or more components in the target standard information acquisition and presentation system 100 (e.g., server 110, storage 120) via the network 140.

Network 140 may facilitate the exchange of information and/or data. In some embodiments, one or more components of the target criteria information acquisition and presentation system 100 (e.g., server 110, storage 120, and user terminal 130) may send information and/or data to other components in the target criteria information acquisition and presentation system 100 via the network 140. For example, server 110 may obtain a user query from user terminal 130 via network 140. For another example, the server 110 may transmit the at least one recommended standard information to the user terminal 130 to cause the user terminal 130 to present the at least one recommended standard information. In some embodiments, the network 140 may be any form of wired or wireless network, or any combination thereof. By way of example only, network 140 may include a cable network, a wired network, a fiber optic network, a telecommunications network, an intranet, the internet, a Local Area Network (LAN), a Wide Area Network (WAN), a Wireless Local Area Network (WLAN), a Metropolitan Area Network (MAN), a Public Switched Telephone Network (PSTN), a bluetooth network, a zigbee network, a Near Field Communication (NFC) network, the like, or any combination of the above. In some embodiments, network 140 may include one or more network access points. For example, network 140 may include wired or wireless network access points, such as base stations and/or internet exchange points, through which one or more components of target standard information acquisition and presentation system 100 may connect to network 140 to exchange data and/or information.

Fig. 2 is a block diagram of an exemplary system according to one or more embodiments of the present description, in some embodiments of which a system for obtaining target criteria information may include a first obtaining module 202, a second obtaining module 204, a first determining module 206, a second determining module 208, and a target determining module 210.

The first obtaining module 202 may be used to obtain the question of the target user.

The second obtaining module 204 may be configured to obtain a trajectory factor of the target user. In some embodiments, the trajectory factor reflects at least one behavior of the target user on one or more service platforms. In some embodiments, the second obtaining module 204 is further configured to obtain one or more behaviors of the target user in one or more service platforms within a set time; generating one or more codes based on the one or more behaviors; and splicing the one or more codes to obtain the track factor.

The first determination module 206 may be configured to determine a first set of criteria information corresponding to the question based on the question and a first preset algorithm.

The second determination module 208 may be configured to determine a first type of information related to the standard information based on the trajectory factor and a second predetermined algorithm.

The targeting module 210 may be configured to determine one or more target criteria information based on the first set of criteria information and the first type of criteria information-related information. In some embodiments, the target determination module 210 is further configured to determine the order of the target criteria information based on at least the first category of information related to criteria information. In some embodiments, the object determination module 210 is further configured to filter a first set of criteria information based on the first type of information associated with criteria information to determine one or more object criteria information. In some embodiments, the targeting module 210 is further configured to determine common criteria information of the first set of criteria information and the second set of criteria information; one or more target criteria information are determined from the common criteria information. In some embodiments, the goal determination module 210 is further configured to: determining standard information; operating the first numerical value and the second numerical value of the common standard information; and determining one or more target standard information from the common standard information based on the operation result.

In some embodiments, the system may further include a target output module 212 for outputting one or more target criteria information.

In some embodiments, the system may further comprise: a third acquisition module and an algorithm optimization module. The third acquisition module is used for acquiring feedback of the user on the one or more target standard information; the algorithm optimization module is used for updating the first preset algorithm and/or the second preset algorithm according to feedback of a user.

It should be understood that the system and its modules shown in FIG. 2 may be implemented in a variety of ways. For example, in some embodiments, the system and its modules may be implemented in hardware, software, or a combination of software and hardware. Wherein the hardware portion may be implemented using dedicated logic; the software portions may be stored in a memory for execution by a suitable instruction execution system, such as a microprocessor or specially designed hardware. Those skilled in the art will appreciate that the methods and systems described above may be implemented using computer executable instructions and/or embodied in processor control code, such code being provided, for example, on a carrier medium such as a diskette, CD-or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The system and its modules of one or more embodiments of the present specification may be implemented not only by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but also by software executed by various types of processors, for example, or by a combination of hardware circuits and software (e.g., firmware).

It should be noted that the above descriptions of the candidate display and determination system and the modules thereof are merely for convenience of description, and do not limit one or more embodiments of the present disclosure to the scope of the illustrated embodiments. It will be appreciated by those skilled in the art that, given the teachings of the present system, any combination of the various modules, or the connection of the constituent subsystems to other modules, or the omission of one or more of the modules, may be made without departing from such teachings. For example, the first obtaining module 202, the second obtaining module 204, the first determining module 206, the second determining module 208, the target determining module 210 and the target output module 212 disclosed in fig. 2 may be different modules in a system, or may be a module that implements the functions of two or more modules described above. In some embodiments, the target output module 212 may also be omitted. In some embodiments, the first obtaining module 202 and the first determining module 206 may be two modules, or one module may have both obtaining and determining functions. For example, each module may share one memory module, and each module may have its own memory module. Such variations are within the scope of one or more embodiments of the present description.

In the present specification, "question of user", and "question of user" are the same meaning. For convenience of description, some parts of this specification will simply refer to "first type information related to standard information" as "first type information", which have the same meaning.

Fig. 3 is an exemplary main flow diagram for target criteria information acquisition, shown in accordance with one or more embodiments of the present description. As shown in fig. 3, a method for acquiring target standard information may include:

step 302, a question of the target user is obtained. In some embodiments, step 302 may be performed by the first acquisition module 202.

In some embodiments, the questions of the target user include, but are not limited to, any combination of one or two or more of questions input by manual means, questions input by voice means, questions input by camera word-taking means, and the like. The manually input question is a question in a text input form, for example, a target user can describe a question that the target user wants to consult with in a customer service dialog box by using a text; the problem of the voice mode input can be that the voice acquisition module acquires voice input, and the voice recognition module converts the problem of the voice form into a problem of a text form; the problem of word-taking mode input of the camera can be the problem of acquiring a character form in an image by an image acquisition module, and the problem of converting the character image form into a text form by a character recognition module, wherein the image can be a character screenshot, other pictures with character contents, or other pictures with specific reference or meaning. For example, a logo picture with characters or letters, and the like, the logo picture is identified by using a camera, and then the problem content can be determined. For another example, a common non-text logo is identified, and product-related information corresponding to the logo can be identified. The presentation mode of the question content can be a complete sentence or an incomplete sentence. The complete sentences can be in any sentence form such as question sentences, statement sentences, exclamation sentences, question reversals and the like, and the complete sentences can also be any irregular sentences such as sick sentences, wrong sentences and the like. The incomplete sentence may be any one or a combination of two or more of sentences, phrases, words, and the like, which have missing major structures.

In some embodiments, the user's question may be received by the user terminal 130 and sent to the server 110 (e.g., the first obtaining module 202) over the network 140. For example, the payment platform service, the problem may be data sent by the user terminal 130 to the server 110 through the network 140 after the user input is completed and confirmed; the problem may also be data that the user terminal 130 sends to the server 110 in real time over the network 140 during the user input. The data sent by the user terminal 130 to the server 110 through the network may be all of the user questions or a part of the user questions, for example, the user questions are processed by text, and keywords in the processed user questions are extracted and sent to the server.

The description will use "network trader credit" as an example to illustrate various processing links in the description, and it should be noted that this example only represents one possible implementation of the description to more clearly describe the processes in the flow, and does not have any limitation on the method or flow described in any embodiment of the description.

For example, the user inputs a piece of voice information as "internet merchant credit repayment" in a voice manner, and the user terminal 130 or the first obtaining module 202 may convert the voice information of "internet merchant credit repayment" into text content "internet merchant credit repayment" through the voice recognition module, so as to process the question asked by the user in the subsequent process. In other embodiments, the user may also enter the text "net loan repayment" directly in the text box as a question of the target user.

Step 304, a first set of criteria information is determined based on the question and a first predetermined algorithm. In some embodiments, step 304 may be performed by the first determination module 206.

The standard information may include a standard question, or may be content information corresponding to the standard question, and the content information may include an answer, or may be information related to the standard question. The standard question is abbreviated as a question mark, is a standardized question sentence, and can reduce ambiguity of the question sentence. For example, "how the network provider is paid" and the like correspond to the question "how the network provider is paid". The content information corresponding to the standard question may include an answer to the quiz. The questions and the answers to the questions have a determined corresponding relationship, and the corresponding mode can be one-to-one correspondence, multiple-to-one correspondence or one-to-multiple correspondence. In some embodiments, the presentation of the standard information to the user may include one or more of text, picture, voice, video, and link. Specifically, the presentation mode is a standard question in a text form and an answer thereof, and the user obtains the answer to solve the question by reading the text content. Or the presentation mode is a picture form, the step points of the operation can be displayed in the picture, and the user obtains the answer for solving the problem by viewing the picture. Or, the presentation mode is a standard question and an answer thereof in a voice form, and the user obtains the answer to solve the question by listening to the voice content. Or the presentation mode is a video mode, the video can completely demonstrate the process of solving the problems of the user, and the user obtains the answers for solving the problems by watching the video. Or, the presentation mode is a link mode, and the user accesses the page recorded with the user question answer by clicking the link to obtain the answer for solving the question.

In some embodiments, determining the first set of criteria information based on the question and the first predetermined algorithm may include performing a correlation query in a database based on the user question to obtain criteria information matching the question. This method will be described in more detail below.

In some embodiments, text processing and/or semantic analysis may be performed based on the question of the target user, and a keyword corresponding to the question may be determined, where the keyword may include one or more of a question word (such as why, how, what, and the like), an action (such as borrowing, returning, registering, logging out, and the like), an entity (such as loan, interest, account number, amount, and the like), and a value (including an amount value, time information, and the like). In some embodiments, the first preset algorithm may be a database query algorithm. After the first obtaining module 302 obtains the question of the target user, the first determining module 306 performs semantic analysis on the question of the user, extracts a keyword in the question of the user as an input of a first preset algorithm or a query algorithm, and then queries a corresponding question and/or answer from a question database. The query algorithm may include any one OR more of a combination of an equivalent query (=), a range query (>, <, betweeen, IN), a fuzzy query (LIKE), an intersection query (AND), a union query (OR). The query algorithm may retrieve corresponding standard information from the standard information database as output by inputting the information. In some embodiments, the criteria information in the criteria information base may have an id, title, or index. In this way, the first set of criteria information may be determined.

For example, when the user inputs a question as "how to loan," after the first obtaining module 302 obtains the user question, the first determining module 306 performs semantic analysis on the question, extracts keywords "how" and "loan" of the question, performs query matching on the keywords in the standard information database through a query algorithm, and determines the standard information as a first set of standard information. For example, query matching is performed in the standard information database, and answers such as "repayment procedure of internet merchant loan", "flower repayment procedure", "how to repay money treasures" and the like are obtained.

In some embodiments, the first preset algorithm may be a processing algorithm implemented by a machine learning model, and correspondingly, determining the first set of criteria information based on the question and the first preset algorithm may include determining criteria information corresponding to the question based on the user question and the machine learning model. In some embodiments, the user's question may be directly used as input to a machine learning model, resulting in standard information or standard information and corresponding probability values corresponding to the question. In some embodiments, the semantic analysis result of the user question may also be used as an input of the machine learning model, and standard information or standard information and a corresponding probability value corresponding to the question are obtained. In some embodiments, the machine learning model may be a Natural Language Processing (NLP) model including, but not limited to, a Bert model, a DSSM model, a CNN-DSSM model, an LSTM-DSSM model, a BCNN model, an ABCNN model, a Hybrid CNN model, and the like. In some embodiments, the first determination module 206 may use the machine learning model to match the user's question with a corresponding first set of criteria information. An embodiment in which the user's question is directly input into the machine learning model to be processed and the first set of standard information is determined will be described below.

In some embodiments, the machine learning model may include a model of one, such as a Bert model. For example only, the model may include a feature extraction layer, a classification layer, and an output layer. Wherein, the feature extraction layer may further include one or more of a word embedding layer, a context expression layer or a vector expression layer. After the first obtaining module 302 obtains the user question, the first determination model 306 inputs the user question into the model. The feature extraction layer of the model can process the user question through word segmentation processing, vector expression (such as word embedding processing) processing and context expression (extracting the context relation of the word segmentation), and obtain the vector expression of each word/word in the user question text after full-text semantic information is fused. In some embodiments, the feature extraction layer may tokenize the input question and may embed the results of the tokenization into a vector representation of a single word and/or phrase (e.g., "cash-out," "repayment"). Context information may then be extracted by convolving any two adjacent vector representations of the at least two vector representations. The convolution information can be maximally pooled to obtain semantic information for the user problem. For example, a vector representation corresponding to each maximum convolution information may be selected. Then, each convolution is subjected to Layer Normalization (Layer Normalization), each vector is subjected to transform encoding (transform Encode) (for example, linear transformation; for example, a vector is transformed into id of int to represent), and finally, the Bert model can transform the input user question into a vector representation fused with full-text semantic information.

And then, a classification layer in the machine learning model classifies the vector representation of the user question to obtain one or more standard information, and finally, an output layer outputs the one or more standard information. In some embodiments, the classification layer may also determine one or more criteria information that match the vector representation of the user question and a first value that reflects a degree of match or probability to be selected by the user. For example, the user inputs the question "how to pay", the first obtaining module 302 inputs the question "how to pay" into a machine learning model trained in advance after obtaining the user question, and the feature extraction layer of the model divides the question "how to pay" into words and converts the divided words into word vectors. The classification layer of the model processes the vectors, and the output layer of the model can output standard information (such as a question) matched with the question of the user and the probability thereof, for example, "repayment flow of network provider loan (probability 0.72)" "which is the way of bei repayment? (probability 0.24) "," how is the payment due to the net trader determined? (probability 0.04) ". And determines the output information as a first set of standard information.

In some embodiments, the machine learning model may also include a plurality of models. For example only, a Word Vector model (e.g., word2Vector, etc.) and a classification model may be included, where the Word Vector model is used to convert a user question into a Vector expression, and the classification model (e.g., masked LM, etc.) is used to process the Vector expression corresponding to the user question, so as to obtain one or more standard information.

It should be noted that, the determining of the first set of standard information based on the question may include determining one or more corresponding questions based on the user question, may also include content information corresponding to the determined one or more questions, and may also be a combination of the one or more questions and the content information thereof. The description is not intended to be limiting in any way.

Step 306, obtaining the track factor of the target user. In some embodiments, step 306 may be performed by the second acquisition module 204.

In some embodiments, the trajectory factor reflects at least one behavior of the target user in one or more service platforms and/or personal attribute information of the target user.

In some embodiments, the trajectory factor reflects an operation behavior of the target user in one or more service platforms, where the service platform may include a platform where the target standard information acquisition and presentation system is located, and may also include other platforms besides the platform where the target standard information acquisition and presentation system is located. For example, the platform where the target standard information acquisition and presentation system is located is a shopping platform, and the other platform may be a financial service platform or the like. The platform where the target standard information acquisition and presentation system is located and the other platforms may have data interaction (for example, the platform a may acquire all or part of the behavior record of the user on the platform B), or may be isolated from each other. In some embodiments, the behavior may include browsing, clicking, registering, downloading, uninstalling, sharing, transferring, cashing, agreeing, paying, and the like, and their operation objects. The operation object may include an accessed web page, newly registered service contents, money type, and the like. The activity may also include parameter information for the activity including, but not limited to, time parameters, value parameters (e.g., transfer amount, cash withdrawal amount, etc.), frequency parameters, etc. For example, after the user transfers money at the service platform, the action includes information of the transfer action, time information of the transfer, amount information of the transfer and the like. In some embodiments, the personal attribute information of the user may include basic information of the user, such as gender, age, academic calendar, native place, occupation, registration time, preference, character tendency or behavior habit, and the like of the user.

For example, the target user registers an account "internet trader credit" in a certain service platform, the second obtaining module 204 obtains the behavior "registration" of the target user, obtains a behavior object "internet trader credit" of the target user, and uses the information as a trajectory factor of the user or determines the trajectory factor based on the information.

In some embodiments, the behavior trace and/or personal attribute information of the user may be obtained by the user terminal 130 and sent to the server 110 (e.g., the second obtaining module 202) via the network 140. The server 110 obtains the track factor of the user according to the behavior track and/or the personal attribute information of the user. In some embodiments, the user's behavioral tracks and/or personal attribute information may also be stored as historical data in a storage device, recalled by the server 110. For more description of obtaining the user trajectory factor, refer to the related description elsewhere in this specification, for example, the related description of fig. 7.

And 308, determining a first type of information related to the standard information based on the track factor and a second preset algorithm. In some embodiments, step 308 may be performed by the second determination module 208.

The first type of information may include screening assistance information or may include a second set of criteria information. Correspondingly, in some embodiments, determining the first type of information based on the trajectory factor and the second preset algorithm may include determining a screening condition based on the trajectory factor and the second preset algorithm; in other embodiments, determining the first type of information based on the trajectory factor and the second preset algorithm may include determining the second type of criteria information based on the trajectory factor and the second preset algorithm. These two embodiments will be described in detail below.

In some embodiments, the first type of information may include screening assistance information reflecting a screening condition of the target standard information. In some embodiments, the screening assistance information or the screening condition may include scoping information, conditionality information, or the like, which plays a limiting role.

In some embodiments, the first set of standard information may be filtered by filtering the auxiliary information or filtering conditions to obtain one or more target standard information. In some embodiments, after the track factor of the target user is obtained, a second preset algorithm performs keyword extraction on the track factor to obtain corresponding entity information, and then determines a screening condition associated with the keyword or the entity information. In some embodiments, a second preset algorithm may classify the trajectory factors to determine a scoping condition for screening.

Still taking "network provider credit" as an example for explanation, for example, the target user registers the "network provider credit" service, and the track factor information corresponding to the target user may include "registration" as an operation behavior and "network provider credit" service as a behavior object. After the system acquires the track factor information, the second preset algorithm can perform relevant processing on the track factor information, extract a keyword 'net trader credit', and then obtain a screening condition 'a problem related to the net trader credit' based on the keyword.

In some embodiments, the first type of information may also include a second set of standard information corresponding to the trajectory factors. In some embodiments, the second set of standard information may include a question related to the trajectory factor of the user, may also include answer information corresponding to the question, and may also include the question and corresponding answer information. Correspondingly, in some embodiments, determining the first type of information based on the trajectory factor and the second preset algorithm may further include determining the second set of standard information based on the trajectory factor and the second preset algorithm. The detailed description about the second set of standard information may refer to the related description of the first set of standard information.

In some embodiments, the second pre-set algorithm may include a machine learning model, which is referred to as a "factor prior model" hereinafter, but should not be named as a limitation on the model, to distinguish it from the machine learning model in the first pre-set algorithm. The "factor prior model" may be a classification model, a regression model, or a model of the same type as the machine learning model associated with the first predetermined algorithm. Correspondingly, in some embodiments, the machine learning model may include a classification model, i.e., a second set of criteria information may be output based on the trajectory factors and the machine learning model; in other embodiments, the machine learning model may include a regression model, i.e., the second set of criteria information and its corresponding probability values may be determined based on the trajectory factors and the machine learning model. These two embodiments will be described separately below.

In some embodiments, the "factor prior model" comprises a classification model. The classification models include, but are not limited to, decision trees, nearest neighbor classifiers, naive bayes classifiers, bayesian Belief Networks (BBNs), neural networks, support Vector Machines (SVMs), and the like. In some embodiments, the acquired trajectory factors are input as input data into a pre-trained machine learning model, and the machine learning model can automatically output a second set of standard information corresponding to the trajectory factors. The acquisition or training process for the machine learning model is described in detail below.

For example, the target user clicks a link related to "promote amount" on the "network provider credit" platform and views related information, the second obtaining module 204 obtains behaviors "click" and "view" of the target user, obtains a behavior object "network provider credit" platform, "promote amount" link and information of the target user, and uses the information as a track factor of the user. And labeling and coding the track factors, converting the track factors into identification id or vector representation, and classifying the identification id or vector representation. A second set of standard information will be determined by the classification model, for example: a second set of questions, "how to promote the credit line of the web merchant? "," what is the upper limit of the net credit? "," payment process for internet trader loan ".

In some embodiments, the "factor prior model" may also include a regression model, which may be a linear regression model, a logistic regression model, a polynomial regression model, a stepwise regression model, a ridge regression model, a lasso regression model, an elastic regression model, or the like. In some embodiments, the "factor prior model" includes, but is not limited to, a Bert model, a DSSM model, a CNN-DSSM model, a LSTM-DSSM model, a BCNN model, an ABCNN model, a Hybrid CNN model, and the like.

In some embodiments, the acquired trajectory factors are input as input data into a machine learning model trained in advance, and the machine learning model can automatically output the second set of standard information and the predicted probability values corresponding to the second set of standard information. The process of obtaining or training the machine learning model is described in detail below.

Step 310, one or more target criteria information is determined based on the first set of criteria information and the first type of information related to criteria information. In some embodiments, step 310 may be performed by the targeting module 210.

In some embodiments, the target criteria information includes more accurate one or more criteria information that is ultimately determined based on information processing of the user's question and the user's trajectory factor. In some embodiments, the finally determined target criteria information may be considered as a combination that will have an order of presentation when presented to the user. In some embodiments, when the finally determined target standard information is one, it may also be regarded as a combination. In some embodiments, the presentation order may include a positional order, may also include a temporal order, and the like. In some embodiments, the order may be determined based on a sequence output by the algorithm, may also be determined based on a matching degree or a click probability of each piece of standard information, and may also be determined randomly.

In some embodiments, the order of the target criterion information may be determined based on the first type of information. In some embodiments, the first type of information includes a filtering condition, items in the first set of standard information are sorted based on the filtering condition, and the first set of standard information arranged in the determined order is determined as the target standard information. For example, the first type of information includes a filter condition "question related to cyber credit", and the first set of standard information includes questions and answers of "flower payment procedure", "payment procedure of cyber credit", "how to pay financial treasures", and the like. Therefore, the bid questions meeting the screening condition can be preferentially displayed according to the first type of information, and the rest bid questions are displayed later. That is, it can be determined that the plurality of target standard information are questions and answers such as "repayment procedure of internet trader loan", "repayment procedure of flower", "how to repay treasures", etc., wherein if the plurality of target standard information are regarded as a combination, the combination has a corresponding sequence, in this embodiment, the repayment procedure of internet trader loan "is in a forward sequence, in other embodiments, the plurality of target standard information may be ranked according to other ranking rules as desired.

In some embodiments, a first set of criteria information may be filtered based on a first type of information to determine one or more target criteria information. By way of example only, the first type of information includes a filtering condition "question related to cyber credit", and the first set of standard information includes questions such as "flower payment flow", "payment flow of cyber credit", "how to pay for financing treasures", and the like, and answers thereof. Therefore, the questions meeting the screening condition can be selected according to the first type of information, and the rest questions are discarded. That is, one target standard information may be determined as "payment flow of network provider credit".

In some embodiments, the first type of information includes a second set of standard information, and correspondingly, determining the target standard information based on the first set of standard information and the first type of information may further include determining common standard information based on the first set of standard information and the second set of standard information, and then determining the target standard information from the common standard information. In some embodiments, the first type of information further includes a second set of standard information and a second value corresponding thereto, the first set of standard information further includes a first value corresponding to the standard information, and correspondingly, determining the target standard information based on the first set of standard information and the first type of information may further include determining the common standard information, then performing an operation on the first value corresponding to the common standard information and the second value, and determining the target standard information based on a result of the operation.

Further description regarding determining target criteria information based on the first set of criteria information and the first type of information may be found elsewhere herein, e.g., in relation to the description of FIG. 4.

At step 312, one or more target criteria information is output. In some embodiments, step 312 may be performed by target output module 212.

In some embodiments, the presentation includes, but is not limited to, text form, voice form, image form, video form, and the like. The output is to output the target standard information to the client so that the user can intuitively receive the target standard information.

In some embodiments, the target criteria information may include criteria questions; content information corresponding to standard questions may also be included; a combination of standard questions and their content information may also be included. In some embodiments, the content information of the standard question may be understood as an answer corresponding to the standard question.

In some embodiments, when outputting the target standard information, one or more pieces of standard information may be also regarded as a combination, and the combination may be output. When the combination contains a plurality of standard information, there is a presentation order. For example, a plurality of questions and content information thereof may be output to the user, the plurality of questions and the content information thereof may be output sequentially in a certain order, or may be arranged in sequence in a certain order, the order may be determined by a probability value of each question, each question may point to an answer text, an answer diagram, or a demonstration video of the question, and is used to provide a method or an answer for solving a corresponding standard question to the user, and the user may view the content of the answer by clicking any one question.

It should be noted that the above description of flowchart 300 is for purposes of example and illustration only and is not intended to limit the applicability of one or more embodiments of the present description. Various modifications and alterations to flow 300 may occur to those skilled in the art, as guided by one or more of the embodiments of the present description. However, such modifications and variations are intended to be within the scope of one or more embodiments of the present disclosure. For example, in some embodiments, step 312 may be omitted. For another example, in some embodiments, the order of steps 302, 304 and steps 306, 308 may be interchanged. Steps 302, 304 and/or 306, 308 may be performed on the same device or may be performed on different devices.

FIG. 4 is a diagrammatic view of a sub-flow for determining target criteria information in accordance with certain embodiments shown in the present description. The process 400 may be performed by the target criteria information acquisition and presentation system 100, and in particular, may be implemented by the targeting module 210. The operation of the process shown below is for illustration purposes only. In some embodiments, process 400, when implemented, may add one or more additional operations not described in one or more embodiments of the specification, and/or delete one or more operations described herein. Additionally, the order in which the process operations are illustrated in FIG. 4 and described below is not intended to be limiting.

In some embodiments, the first type of information includes a second set of standard information including one or more standard questions and/or their corresponding content information. In some embodiments, the second set of criteria information intersects the first set of criteria information. The intersection set here may be understood as that the second set of standard information has the same or similar standard information as the first set of standard information.

In some embodiments, the similarity may be understood as meaning that the semantic distance is less than a set threshold, or that the two or more criterion information are similar based on a semantic distance algorithm or a text matching algorithm. In some embodiments, the same or similar criteria information may include exactly the same criteria question; in some embodiments, the same or similar criteria information may include criteria questions of higher similarity. The similarity of the standard question can be set in advance based on specific situations, for example, the similarity is set according to the text contact degree in the standard question, for example, several same characters or words exist; or what the same field is in the whole standard problem. In some embodiments, the same or similar criteria information may include exactly the same answer content; in some embodiments, similar answer content may also be included. In some implementations, similar answer content can include answer content with higher text overlap. In step 402, common standard information in the first set of standard information and the second set of standard information is determined. In some embodiments, the common criteria information may be an intersection of the first set of criteria information and the second set of criteria information, where an understanding of the intersection may be with reference to the above paragraph. In some embodiments, the common criteria information may further include the same or similar criteria information after the first and second sets of criteria information are filtered, respectively. The screening may include screening based on a preset rule. The public standard information may include the quiz, the answer corresponding to the quiz, or a combination of the two. In some embodiments, the determining common standard information in the first set of standard information and the second set of standard information may be determined based on an equivalence match, an intersection query, or the like. For example, the first group of standard information includes the questions and answers of ' payment procedure of flower ', ' payment procedure of internet merchant loan ', ' how to pay money for treasury ', etc., and the first group of standard information includes ' how to promote the line of internet merchant loan? "," what is the upper limit of the net credit? "," the payment flow of the online merchant loan "," how long the payment time of the online merchant loan is determined? "the answers to the questions and answers can be determined to include" the payment flow of the online loan "," how is the payment time of the online loan determined? ".

In some embodiments, the options in the intersection may be used as the common standard information, and a part of the options in the intersection may also be selected as the common standard information, where the selection condition may be the same or similar degree. For example, in the above-mentioned intersection, "the payment flow of the cyber loan" is more certain than "how is the payment time of the cyber loan? "is close, so the" payment flow of the network trader loan "is used as the public standard information.

In some embodiments, the first set of criteria information includes a first numerical value corresponding to each criteria information thereof, and the second set of criteria information includes a second numerical value corresponding to each criteria information thereof. In some embodiments, the first numerical value of the first set of criteria information may be a probability value, a match value, or a weight value. The second numerical value of the second set of standard information may also be a probability value, a matching value, or a weight value, and these numerical values may be probabilities of the respective standard information calculated by a preset algorithm. The preset algorithm may be a first preset algorithm or a second preset algorithm.

And step 404, operating the first numerical value and the second numerical value of the common standard information. The operation may be addition or multiplication. For example, the common standard information includes "the payment flow of the internet trader loan", "how is the payment time of the internet trader loan determined? "two questions, wherein" the repayment process of the network loan "has a first value of 0.9 from the first set of standard information and a second value of 0.7 from the second set of standard information, and the first value and the second value are multiplied to obtain a target value of 0.63 for the common standard information. "how is the payment time for the network merchant credited determined? "the first value from the first set of standard information is 0.6, the second value from the second set of standard information is 0.8, and the first value and the second value are multiplied to obtain the target value of the common standard information which is 0.48.

And 406, determining one or more target standard information from the common standard information based on the operation result. In some embodiments, the ranking may be performed based on the target value of the common standard information, and the standard information of the top several bits (e.g., the top 3 bits) is recommended as the target standard information from the ranking result. For example, "the payment flow of the network provider credit", "how long is the payment time of the network provider credit? "two questions and answers thereof are the one or more target criterion information. In some embodiments, a threshold (e.g., 0.5) may also be set for the target value of the common standard information, and when the target value of the common standard information is greater than the threshold, the common standard information is determined as the target standard information. For example, the question "the payment process of the internet trader loan" and the answer thereof can be used as the one or more target standard information.

FIG. 5 is a diagrammatic illustration of a training flow of a machine learning model in accordance with a first pre-set algorithm as shown in some embodiments of the present description. The operation of the process shown below is for illustration purposes only. In some embodiments, process 500 may be implemented with one or more additional operations not described in one or more embodiments herein, and/or with one or more operations described herein. Additionally, the order in which the process operations are illustrated in FIG. 5 and described below is not intended to be limiting.

Step 502, the questions of the historical users and the corresponding standard information are obtained.

The historical user may include user data on the service platform in the history. In some embodiments, the questions of the historical user may be determined based on historical user data, including but not limited to questions entered manually, questions entered by voice, questions entered by camera-word, and the like, in any combination of one or more than two. The corresponding standard information may include standard information clicked by a historical user after inputting data; the method can also comprise the steps of historical user approval standard information after data input; the method can also comprise standard information of the maximum browsing times of the historical user after data input; and the standard information which is shared and/or forwarded most times by historical users after data input can also be included. The standard information may include a question, an answer corresponding to the question, or a combination of both.

At step 504, an initial first machine learning model is trained. The machine learning model may be a Natural Language Processing (NLP) model including, but not limited to, a Bert model, a DSSM model, a CNN-DSSM model, an LSTM-DSSM model, a BCNN model, an ABCNN model, a Hybrid CNN model, and the like.

In some embodiments, the first machine learning model may be trained by: acquiring a first training sample set; the first training sample set includes the input questions of the historical user and the standard information corresponding to the input questions, that is, the questions of the historical user obtained in step 502 and the standard information corresponding to the questions. In some embodiments, the standard information corresponding to the input question may include a standard question and/or a corresponding answer selected by a click after a historical user input question; standard questions and/or corresponding answers that historical users complied with after entering questions may also be included; it may also include historical standard questions and/or corresponding answers that the user forwarded after entering the question. And training an initial first machine learning model by using the first training sample set to obtain a trained first machine learning model.

Step 506, a trained first machine learning model is obtained. The trained first machine learning model may be responsive to a user question, and in some embodiments, the trained first machine learning model may be capable of outputting one or more corresponding criteria information based on the user question or a corresponding word vector of the user question to determine a first set of criteria information. In some embodiments, after the first obtaining module 202 obtains the user question, at least one first set of criteria information may be determined using the first machine learning model. The first set of standard information at least comprises one of a question, an answer corresponding to the question or a combination thereof. With respect to the method of using the first machine learning model, more detailed descriptions can be found elsewhere herein.

In some embodiments, the model parameters may be adjusted back according to the difference between the predicted output of the model (e.g., predicted criteria information) and the reference criteria for the purpose of training or optimizing the model. In some embodiments, the model may be optimized or updated by adding sample data. In some embodiments, when the difference satisfies a predetermined condition, for example, the predicted accuracy of the model is greater than a predetermined accuracy threshold, or the value of the Loss Function (Loss Function) is less than a predetermined value, the training process is stopped, i.e. the trained first machine learning model is obtained.

In some embodiments, the difference between the output value and the reference value may be determined according to the feedback of the user, and the optimization algorithm may be performed by increasing sample data. In some embodiments, the user's feedback may include, but is not limited to, the user clicks and/or number of clicks, browsing and/or browsing times, forwarding and/or forwarding times, sharing and/or number of shares, giving up viewing, and the like.

FIG. 6 is a diagrammatic illustration of a training flow of a machine learning model in accordance with a second pre-set algorithm as shown in some embodiments of the present description. The operation of the process shown below is for illustration purposes only. The process 600 may be performed by the second determination module 208 of the target criteria information acquisition and presentation system 100. The operation of the process shown below is for illustration purposes only. In some embodiments, process 600 may be implemented with one or more additional operations not described in one or more embodiments herein, and/or with one or more deletions described herein. Additionally, the order in which the process operations are illustrated in FIG. 6 and described below is not intended to be limiting.

In some embodiments, step 308 determines that the first type of information related to the standard information can be automatically processed by a machine algorithm based on the trajectory factor and a second preset algorithm; in some embodiments, the machine algorithm may be processed by a machine learning model, resulting in fast processing speed and high accuracy.

Step 602, obtaining the track factor of the historical user and the first type of information related to the standard information corresponding to the track factor. The trajectory factor reflects at least one behavior of the historical user in one or more service platforms, and the historical user may include user data already on the service platform. In some embodiments, the trajectory factor also reflects personal attribute information of the historical user. In some embodiments, the first type of information related to the standard information corresponding thereto or the first type of information corresponding thereto may include one or more standard information corresponding to the trajectory factor of the historical user, and may further include other auxiliary information or filtering conditions related to the trajectory factor of the historical user, and the filtering conditions may include scoping information, conditional information, and the like.

At step 604, an initial second machine learning model is trained. The second machine learning model may include a model similar to the first machine learning model. For example, the second machine learning model includes, but is not limited to, a Bert model.

In some embodiments, said training an initial second machine learning model comprises: acquiring a second training sample set; and training an initial second machine learning model by using the second training sample set to obtain a trained second machine learning model. The second training sample set includes the trajectory factors of the historical users obtained in step 602 and the first type information corresponding to the trajectory factors.

In some embodiments, the second machine learning model may be trained using the trajectory factors in the sample set as input data and the first type of information corresponding to the trajectory factors as output data or reference criteria.

And 606, obtaining a trained second machine learning model. The trained second machine learning model may be responsive to the trajectory factor of the target user, and in some embodiments, the trained second machine learning model may be capable of determining a screening condition or standard information corresponding to the trajectory factor based on the trajectory factor of the target user to obtain the determined first type of information. In some embodiments, after the second obtaining module 204 obtains the user trajectory factor, at least one first type of information related to the standard information may be determined using a second machine learning model. In some embodiments, the first type of information related to the standard information or the first type of information includes at least one of a quiz, an answer corresponding to the quiz, or a combination thereof; in some embodiments, the first type of information may further include screening assistance information reflecting a screening condition of the target standard information. In some embodiments, the filtering condition can filter the first set of standard information to serve the target standard information. With respect to the method of use of the second machine learning model, more detailed descriptions may be found elsewhere herein.

In some embodiments, the model parameters may be adjusted back according to differences between the predicted output of the model (e.g., predicted criteria information) and the reference criteria for the purpose of training or optimizing the model. In some embodiments, the model may be optimized or updated by adding sample data. In some embodiments, when the difference satisfies a predetermined condition, for example, the predicted accuracy of the model is greater than a predetermined accuracy threshold, or the value of the Loss Function (Loss Function) is less than a predetermined value, the training process is stopped, i.e., the trained second machine learning model is obtained.

In some embodiments, the model parameters may also be adjusted by obtaining user feedback on the one or more target criteria information, so as to achieve the purpose of training or optimizing the model. The feedback includes, but is not limited to, user clicks and/or click times, browsing and/or browsing times, forwarding and/or forwarding times, sharing and/or sharing times, giving up viewing, and the like.

In some embodiments, user feedback on the one or more target criteria information may also be obtained; and updating the first preset algorithm and/or the second preset algorithm according to the feedback of the user. The user's feedback may include, but is not limited to, the user clicks and/or times, browsing and/or browsing times, forwarding and/or forwarding times, sharing and/or sharing times, giving up viewing, and the like. The first preset algorithm and/or the second preset algorithm may affect the order of the target standard information. In some embodiments, updating the first preset algorithm and/or the second preset algorithm includes collecting a new training sample, and retraining the machine learning model associated with the first preset algorithm and/or the second preset algorithm to improve the output accuracy thereof. In some embodiments, updating the first preset algorithm and/or the second preset algorithm includes adjusting the filtering condition of the standard information (e.g., optimizing the filtering range), and improving the accuracy of the filtering. In some embodiments, updating the first preset algorithm and/or the second preset algorithm includes adjusting a threshold corresponding to the first value and the second value to improve the output accuracy. In some embodiments, updating the first preset algorithm and/or the second preset algorithm further includes adjusting a logic or method of extracting keywords from the user question to improve accuracy of keyword acquisition.

FIG. 7 is a sub-flow diagram illustration shown in some embodiments in FIG. 3 of obtaining a trajectory factor for a target user. The process 900 may be performed by the target criteria information acquisition and presentation system 100, and may be specifically implemented by the second acquisition module 204. The operation of the process shown below is for illustration purposes only. In some embodiments, process 700, when implemented, may add one or more additional operations not described in one or more embodiments of the present specification, and/or delete one or more operations described herein. Additionally, the order in which the process operations are illustrated in FIG. 7 and described below is not intended to be limiting.

In some embodiments, one or more behaviors of the user in the service platform within a set time may be manually collected and input into the system; or may be machine acquired and encoded. The following embodiment will describe a process of acquiring and encoding user behaviors through a machine.

Step 702, acquiring one or more behaviors of a target user in a service platform within a set time. In some embodiments, the service platform may be a platform where the target standard information acquisition and presentation system is located, or may be an associated platform that is associated with the platform where the target standard information acquisition and presentation system is located and that is capable of or has a right to acquire related information. In some embodiments, the action may be an operation that any user may perform on the service platform, such as browsing, clicking, registering, downloading, uninstalling, sharing, transferring, withdrawing, agreeing, paying, and the like. The behavior may also include parameter information of the behavior, including but not limited to time parameters, numerical parameters, frequency parameters, and the like. The set time may be a certain time after the user completes the action, for example, within one week, within three days, within one day, within five minutes, and so on.

Step 704, generating one or more codes based on the one or more behaviors. In some embodiments, the encoding may be encoding the user behavior trace, and the encoding is for labeling and normalization. The coding method can adopt different methods according to different services. For example, if a user accesses a net friend link, the link is encoded as id of int to represent the link; similar processing is carried out on some continuous numerical values, for example, the transfer amount is divided into five levels of 0-9.99, 10-49.99, 50-199.99, 200-999.99, 1000 and more, and the levels are also converted into id of int to represent the transfer amount.

Step 706, concatenating the one or more codes to obtain the trajectory factor. The trajectory factor is obtained based on one or more code splices corresponding to the one or more operation behaviors.

In some embodiments, the behavior of the target user may also include personal attribute information of the target user in some embodiments. The personal attribute information may include basic information, characters, preferences, and the like. Correspondingly, in some embodiments, it is also possible to encode the user personal attribute information to obtain the corresponding trajectory factor. For example, after the consumption preference of the user is obtained, the preference of the user is coded as id of int to represent; the academic information of the user is also encoded into id of int to represent; the user's revenue information (if available) is also encoded as an id for int to represent.

In some embodiments, the user's behavior and personal information may also be synthetically encoded to obtain corresponding trajectory factors. For example: the loan behavior of the user who likes installment is coded as id of int to represent; the loan behavior of the user with the annual income of 15-20 ten thousand is also converted into id of int to represent.

Benefits that may be realized by one or more embodiments of the present description include, but are not limited to: in the process of determining and outputting the target standard information, one or more embodiments of the present disclosure can greatly improve the recognition capability, especially improve the recognition rate of the fuzzy problem, by means of the behavior track information of the user. The method and the system can improve the accuracy of the customer service question-answering system in answering the user questions, convert the original 'unrecognizable' information into 'recognizable' information, and greatly improve the ability of recognizing the question of the user. It is to be noted that different embodiments may produce different advantages, and in different embodiments, any one or combination of the above advantages may be produced, or any other advantages may be obtained.

Having thus described the basic concept, it will be apparent to those skilled in the art that the foregoing detailed disclosure is to be considered merely illustrative and not restrictive of the scope of one or more embodiments of the present specification. Various modifications, improvements, and adaptations to one or more embodiments of the present disclosure may occur to those skilled in the art, although not explicitly described herein. Such alterations, modifications, and improvements are intended to be suggested in one or more embodiments of this disclosure, and are intended to be within the spirit and scope of the exemplary embodiments of one or more embodiments of this disclosure.

Also, the use of specific language in describing one or more embodiments of the specification has been used to describe the embodiments of the one or more embodiments of the specification. Reference to "one embodiment," "an embodiment," and/or "some embodiments" means a feature, structure, or characteristic described in connection with at least one embodiment of one or more embodiments of the specification. Therefore, it is emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, some features, structures, or characteristics of one or more embodiments of the specification may be combined as appropriate.

Moreover, those skilled in the art will appreciate that aspects of one or more embodiments of the present description may be illustrated and described in terms of several patentable species or contexts, including any new and useful combination of procedures, machines, articles, or materials, or any new and useful improvement thereof. Accordingly, aspects of one or more embodiments of the present description may be performed entirely by hardware, entirely by software (including firmware, resident software, micro-code, etc.), or by a combination of hardware and software. The above hardware or software may be referred to as "data block," module, "" engine, "" unit, "" component, "or" system. Furthermore, aspects of one or more embodiments of the present description may be presented as a computer product, including computer readable program code, embodied in one or more computer readable media.

The computer storage medium may comprise a propagated data signal with the computer program code embodied therewith, for example, on baseband or as part of a carrier wave. The propagated signal may take any of a variety of forms, including electromagnetic, optical, etc., or any suitable combination. A computer storage medium may be any computer-readable medium that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code located on a computer storage medium may be propagated over any suitable medium, including radio, cable, fiber optic cable, RF, or the like, or any combination of the preceding.

Computer program code required for operation of portions of one or more embodiments of the present description may be written in any one or more programming languages, including an object oriented programming language such as Java, scala, smalltalk, eiffel, JADE, emerald, C + +, C #, VB.NET, python, and the like, conventional programming languages, such as C, visualBasic, fortran2003, perl, COBOL2002, PHP, ABAP, dynamic programming languages, such as Python, ruby, and Groovy, or other programming languages, and the like. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or processing device. In the latter scenario, the remote computer may be connected to the user's computer through any network format, such as a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet), or in a cloud computing environment, or as a service, such as a software as a service (SaaS).

Additionally, the order in which elements and sequences are described, the use of numerical letters, or the use of other designations in one or more embodiments of the present description is not intended to limit the order in which processes and methods are described in one or more embodiments of the present description, unless explicitly recited in the claims. While certain presently contemplated useful embodiments have been discussed in the foregoing disclosure by way of examples, it is to be understood that such detail is solely for that purpose and that the appended claims are not limited to the disclosed embodiments, but, on the contrary, are intended to cover all modifications and equivalent arrangements that are within the spirit and scope of one or more embodiments herein. For example, although the system components described above may be implemented by hardware devices, they may also be implemented by software-only solutions, such as installing the described system on an existing processing device or mobile device.

Similarly, it should be noted that in the foregoing description of one or more embodiments of the specification, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the embodiments. This method of disclosure, however, is not intended to imply that more features than are expressly recited in one or more of the embodiments in the specification are required. Indeed, the embodiments may be characterized as having less than all of the features of a single embodiment disclosed above.

Numerals describing the number of components, attributes, etc. are used in some embodiments, it being understood that such numerals used in the description of the embodiments are modified in some instances by the use of the modifier "about", "approximately" or "substantially". Unless otherwise indicated, "about", "approximately" or "substantially" indicates that the number allows a variation of ± 20%. Accordingly, in some embodiments, the numerical parameters set forth in the specification and claims are approximations that may vary depending upon the desired properties sought to be obtained by a particular embodiment. In some embodiments, the numerical parameter should take into account the specified significant digits and employ a general digit preserving approach. Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the range in one or more embodiments of the present specification are approximations, in specific embodiments, such numerical values are set forth as precisely as possible within the scope of the application.

Each patent, patent application publication, and other material, such as articles, books, specifications, publications, documents, and the like, cited in connection with one or more embodiments of the specification is hereby incorporated by reference in its entirety. Except where the application history document does not conform to or conflict with the contents of one or more embodiments of the present specification, it is clear that the claims hereof are to be accorded the widest scope limited only to documents that are currently or later appended to one or more embodiments of the present specification. It is noted that the descriptions, definitions and/or use of terms in one or more embodiments of the present specification shall control if the descriptions, definitions and/or use of terms in one or more embodiments of the present specification are inconsistent or contrary to the statements and/or use of terms in one or more embodiments of the present specification.

Finally, it should be understood that the embodiments described in one or more embodiments of this specification are intended only to illustrate the principles of one or more embodiments of this specification. Other variations are possible within the scope of one or more embodiments of the present description. Thus, by way of example, and not limitation, alternative configurations of one or more embodiments of the specification can be viewed as being consistent with the teachings of one or more embodiments of the specification. Accordingly, the embodiments of the present specification are not limited to only those embodiments specifically illustrated and described for one or more of the embodiments of the present specification.

Claims

1. A method for obtaining target criteria information, the method comprising:

obtaining a question of a target user;

determining a first set of standard information based on the question and a first preset algorithm;

acquiring a track factor of the target user; the trajectory factor reflects at least one behavior of the target user on one or more service platforms;

determining first type of information related to standard information based on the track factor and a second preset algorithm;

one or more target criteria information are determined based on the first set of criteria information and the first type of criteria information-related information.

2. The method of claim 1, further comprising outputting one or more target criteria information.

3. The method of claim 1, further comprising:

obtaining feedback of the user on the one or more target standard information;

and updating the first preset algorithm and/or the second preset algorithm according to the feedback of the user.

4. The method according to claim 1, characterized in that the first preset algorithm and/or the second preset algorithm comprises a machine learning model comprising a Bert model.

5. The method of claim 1, wherein determining one or more target criteria information based on the first set of criteria information and the first type of criteria information-related information comprises:

screening a first group of standard information based on the first type of information related to the standard information to determine one or more target standard information; the first type of information related to the standard information includes screening auxiliary information reflecting a screening condition of the target standard information.

6. The method of claim 1, wherein the first type of information related to standard information comprises a second set of standard information; the second set of standard information includes one or more standard information and/or content information corresponding thereto.

7. The method of claim 6, wherein determining one or more target criteria information based on the first set of criteria information and the first type of annotation information comprises:

determining common standard information of the first group of standard information and the second group of standard information;

one or more target criteria information is determined from the common criteria information.

8. The method of claim 6, wherein the first set of criteria information further includes a first numerical value corresponding to each of the criteria information thereof; the second set of criteria information further includes a second value corresponding to each of the criteria information.

9. The method of claim 8, wherein determining one or more target criteria information based on the first set of criteria information and the first type of annotation information-related information comprises:

operating the first numerical value and the second numerical value of the common standard information;

and determining one or more target standard information from the common standard information based on the operation result.

10. The method of claim 9, wherein the operation comprises a multiplication or an addition.

11. A system for obtaining target criteria information, the system comprising:

the first acquisition module is used for acquiring the question of the target user;

the first determining module is used for determining a first group of standard information based on the problem and a first preset algorithm;

the second acquisition module is used for acquiring the track factor of the target user; the trajectory factor reflects at least one behavior of the target user on one or more service platforms;

the second determining module is used for determining the first type of information related to the standard information based on the track factor and a second preset algorithm;

a target determination module that determines one or more target criteria information based on the first set of criteria information and the first type of criteria information-related information.

12. The system of claim 11, further comprising:

and the target output module is used for outputting one or more pieces of target standard information.

13. The system of claim 11, further comprising:

the third acquisition module is used for acquiring feedback of the user on the one or more target standard information;

and the algorithm optimization module is used for updating the first preset algorithm and/or the second preset algorithm according to the feedback of the user.

14. The system according to claim 11, wherein the first preset algorithm and/or the second preset algorithm comprises a machine learning model, the machine learning model comprising a Bert model.

15. The system of claim 11, wherein the goal determination module is further configured to filter a first set of criteria information based on the first type of information related to criteria information to determine one or more target criteria information; the first type of information related to the standard information includes screening auxiliary information reflecting a screening condition of the target standard information.

16. The system of claim 11, wherein the first type of information associated with standard information comprises a second set of standard information; the second set of criteria information includes one or more criteria information and/or its corresponding content information.

17. The system of claim 16, wherein the target determination module is further configured to determine common criteria information of the first set of criteria information and the second set of criteria information;

18. The system of claim 16, wherein the first set of criteria information further comprises a first numerical value corresponding to each criteria information thereof; the second set of criteria information further includes a second value corresponding to each of the criteria information.

19. The system of claim 18, wherein the goal determination module is further configured to: determining common standard information of the first group of standard information and the second group of standard information;

20. The system of claim 19, wherein the operation comprises a multiplication or an addition.

21. An apparatus for obtaining target criteria information, the apparatus comprising at least one processor and at least one memory;

the at least one memory is for storing computer instructions;

the at least one processor is configured to execute at least some of the computer instructions to perform the operations of any one of claims 1-10.