Text processing method and device, electronic equipment and storage medium

Abstract

The present disclosure provides a text processing method and device, electronic equipment and storage medium, relates to the field of artificial intelligence, and in particular to the field of natural language processing based on artificial intelligence. The implementation scheme is: obtaining a to-be-processed text, the to-be-processed text including a question and answer sentence between two objects; obtaining at least one structured field corresponding to the to-be-processed text from a structured field set, each structured field in the structured field set corresponds to a question, and the question corresponds to multiple answers; for each structured field in the at least one structured field, obtaining a structured processing result corresponding to the structured field of the to-be-processed text; and obtaining structured information of the to-be-processed text based on the structured processing result corresponding to each structured field in the at least one structured field of the to-be-processed text.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and in particular to natural language processing based on artificial intelligence, specifically to a text processing method, apparatus, electronic device, computer-readable storage medium, and computer program product. Background Technology

[0002] Artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies mainly include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0003] Artificial intelligence-based natural language processing (NLP) technology analyzes and extracts semantics from language text to understand it, and has been applied in a wide range of fields. For example, in the medical field, by analyzing the dialogue between voice customer service representatives and patients during voice follow-ups, patient information can be processed to obtain the information needed by doctors. This allows doctors to obtain information about patients' conditions or recovery progress without having to conduct in-person follow-ups, thus improving doctors' work efficiency.

[0004] The methods described in this section are not necessarily methods that had been previously conceived or adopted. Unless otherwise specified, no method described in this section should be assumed to be prior art simply because it is included in this section. Similarly, unless otherwise specified, the issues mentioned in this section should not be considered to be accepted in any prior art. Summary of the Invention

[0005] This disclosure provides a text processing method, apparatus, electronic device, computer-readable storage medium, and computer program product.

[0006] According to one aspect of this disclosure, a text processing method is provided, comprising: obtaining a text to be processed, the text to be processed including a question-and-answer statement between two objects; obtaining at least one structured field corresponding to the text to be processed from a set of structured fields, each structured field in the set of structured fields corresponding to a question and the question corresponding to multiple answers; obtaining a structured processing result of the text to be processed corresponding to each of the at least one structured field, the structured processing result indicating the answer in the multiple answers corresponding to the question-and-answer statement of the text to be processed for the question corresponding to the structured field; and obtaining structured information of the text to be processed based on the structured processing result of the text to be processed corresponding to each of the at least one structured field.

[0007] According to another aspect of this disclosure, a text processing apparatus is provided, comprising: a text-to-process acquisition unit configured to acquire text to be processed, the text to be processed including a question-and-answer statement between two objects; a structured field acquisition unit configured to acquire at least one structured field corresponding to the text to be processed from a set of structured fields, each structured field in the set of structured fields corresponding to a question, and the question corresponding to multiple answers; a structured processing result acquisition unit configured to acquire a structured processing result of the text to be processed corresponding to each of the at least one structured field, the structured processing result indicating an answer among the multiple answers corresponding to the question-and-answer statement of the text to be processed for the question corresponding to the structured field; and a structured information acquisition unit configured to acquire structured information of the text to be processed based on the structured processing result of the text to be processed corresponding to each of the at least one structured field.

[0008] According to another aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method according to embodiments of this disclosure.

[0009] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described according to embodiments of this disclosure.

[0010] According to another aspect of this disclosure, a computer program product is provided, including a computer program, wherein the computer program, when executed by a processor, implements the method described according to embodiments of this disclosure.

[0011] According to one or more embodiments of this disclosure, the accuracy of structured information extracted from text containing question-and-answer statements between two objects can be improved.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0013] The accompanying drawings exemplify embodiments and form part of the specification, serving together with the textual description to explain exemplary implementations of the embodiments. The illustrated embodiments are for illustrative purposes only and do not limit the scope of the claims. Throughout the drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0014] Figure 1 A schematic diagram of an exemplary system in which the various methods described herein may be implemented according to embodiments of the present disclosure is shown;

[0015] Figure 2 A flowchart of a text processing method according to an embodiment of the present disclosure is shown;

[0016] Figure 3 A flowchart is shown illustrating the process of obtaining a structured processing result of a text to be processed corresponding to each of at least one structured field in a text processing method according to an embodiment of the present disclosure;

[0017] Figure 4 A schematic diagram of a dual-tower model involved in a text processing method according to an embodiment of the present disclosure is shown;

[0018] Figure 5 A flowchart is shown illustrating the process of obtaining a structured processing result of a text to be processed corresponding to each of at least one structured field in a text processing method according to an embodiment of the present disclosure;

[0019] Figure 6 A flowchart is shown illustrating the process of obtaining structured information of a text to be processed based on the structured processing result of each of at least one structured field in a text processing method according to an embodiment of the present disclosure.

[0020] Figure 7A flowchart of a text processing method according to an embodiment of the present disclosure is shown;

[0021] Figure 8 A structural block diagram of a text processing apparatus according to embodiments of the present disclosure is shown; and

[0022] Figure 9 A structural block diagram of an exemplary electronic device that can be used to implement embodiments of the present disclosure is shown. Detailed Implementation

[0023] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0024] In this disclosure, unless otherwise stated, the use of terms such as "first," "second," etc., to describe various elements is not intended to limit the positional, temporal, or importance relationships of these elements; such terms are merely used to distinguish one element from another. In some examples, the first element and the second element may refer to the same instance of that element, while in other cases, based on the context, they may refer to different instances.

[0025] The terminology used in the description of the various examples described in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context explicitly indicates otherwise, an element may be one or more unless the number of elements is specifically limited. Furthermore, the term "and / or" as used in this disclosure covers any one of the listed items and all possible combinations thereof.

[0026] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0027] Figure 1 A schematic diagram of an exemplary system 100 in which the various methods and apparatus described herein can be implemented according to embodiments of this disclosure is shown. Reference Figure 1 The system 100 includes one or more client devices 101, 102, 103, 104, 105 and 106, a server 120, and one or more communication networks 110 coupling the one or more client devices to the server 120. The client devices 101, 102, 103, 104, 105 and 106 can be configured to execute one or more applications.

[0028] In embodiments of this disclosure, server 120 may run one or more services or software applications that enable text processing methods to be performed.

[0029] In some embodiments, server 120 may also provide other services or software applications, which may include non-virtual and virtual environments. In some embodiments, these services may be provided as web-based services or cloud services, such as to users of client devices 101, 102, 103, 104, 105, and / or 106 under a Software as a Service (SaaS) model.

[0030] exist Figure 1 In the configuration shown, server 120 may include one or more components that implement the functions performed by server 120. These components may include software components, hardware components, or combinations thereof that can be executed by one or more processors. Users operating client devices 101, 102, 103, 104, 105, and / or 106 can sequentially interact with server 120 using one or more client applications to utilize the services provided by these components. It should be understood that various different system configurations are possible and may differ from system 100. Therefore, Figure 1 This is an example of a system used to implement the various methods described herein, and is not intended to be limiting.

[0031] Users can use client devices 101, 102, 103, 104, 105 and / or 106 to receive structured information obtained according to the text processing method of this disclosure. The client devices can provide an interface enabling users of the client devices to interact with them. The client devices can also output information to the user via this interface. Although Figure 1 Only six client devices are described, but those skilled in the art will understand that this disclosure can support any number of client devices.

[0032] Client devices 101, 102, 103, 104, 105, and / or 106 may include various types of computer devices, such as portable handheld devices, general-purpose computers (such as personal computers and laptops), workstation computers, wearable devices, smart screen devices, self-service terminal devices, service robots, gaming systems, thin clients, various messaging devices, sensors, or other sensing devices. These computer devices can run various types and versions of software applications and operating systems, such as Microsoft Windows, Apple iOS, UNIX-like operating systems, Linux or Linux-like operating systems (such as Google Chrome OS); or include various mobile operating systems, such as Microsoft Windows Mobile OS, iOS, Windows Phone, and Android. Portable handheld devices may include cellular phones, smartphones, tablets, personal digital assistants (PDAs), etc. Wearable devices may include head-mounted displays (such as smart glasses) and other devices. Gaming systems may include various handheld gaming devices, internet-enabled gaming devices, etc. Client devices are capable of executing various applications, such as various internet-related applications, communication applications (such as email applications), short message service (SMS) applications, and can use various communication protocols.

[0033] Network 110 can be any type of network well known to those skilled in the art, and can use any of a variety of available protocols (including but not limited to TCP / IP, SNA, IPX, etc.) to support data communication. By way of example only, one or more networks 110 can be a local area network (LAN), an Ethernet-based network, a token ring network, a wide area network (WAN), the Internet, a virtual network, a virtual private network (VPN), an intranet, an extranet, a blockchain network, a public switched telephone network (PSTN), an infrared network, a wireless network (e.g., Bluetooth, WIFI), and / or any combination of these and / or other networks.

[0034] Server 120 may include one or more general-purpose computers, special-purpose server computers (e.g., PC (personal computer) servers, UNIX servers, mid-range servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and / or combination. Server 120 may include one or more virtual machines running a virtual operating system, or other computing architectures involving virtualization (e.g., one or more flexible pools of logical storage devices that can be virtualized to maintain virtual storage devices for servers). In various embodiments, server 120 may run one or more services or software applications that provide the functionality described below.

[0035] The computing unit in server 120 can run one or more operating systems, including any of the aforementioned operating systems and any commercially available server operating system. Server 120 can also run any of a variety of additional server applications and / or middleware applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, etc.

[0036] In some implementations, server 120 may include one or more applications to analyze and merge data feeds and / or event updates received from users of client devices 101, 102, 103, 104, 105 and / or 106. Server 120 may also include one or more applications to display data feeds and / or real-time events via one or more display devices of client devices 101, 102, 103, 104, 105 and / or 106.

[0037] In some implementations, server 120 can be a server for a distributed system or a server integrated with blockchain. Server 120 can also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology. A cloud server is a host product in the cloud computing service system, designed to address the shortcomings of traditional physical hosts and Virtual Private Server (VPS) services, such as high management difficulty and weak business scalability.

[0038] System 100 may also include one or more databases 130. In some embodiments, these databases may be used to store data and other information. For example, one or more of the databases 130 may be used to store information such as audio files and video files. Databases 130 may reside in various locations. For example, a database used by server 120 may be local to server 120, or it may be located away from server 120 and may communicate with server 120 via a network-based or dedicated connection. Databases 130 may be of different types. In some embodiments, the database used by server 120 may be, for example, a relational database. One or more of these databases may store, update, and retrieve data from and from the databases in response to commands.

[0039] In some embodiments, one or more of the databases 130 may also be used by an application to store application data. The databases used by the application may be of different types, such as key-value stores, object stores, or regular stores supported by a file system.

[0040] Figure 1The system 100 can be configured and operated in various ways to enable the application of the various methods and apparatus described in this disclosure.

[0041] According to one aspect of this disclosure, a text processing method is provided. For example... Figure 2 As shown, a text processing method 200 according to some embodiments of the present disclosure includes:

[0042] Step S210: Obtain the text to be processed, which includes question and answer statements between two objects;

[0043] Step S220: Obtain at least one structured field corresponding to the text to be processed from the structured field set, wherein each structured field in the structured field set corresponds to a question and the question corresponds to multiple answers;

[0044] Step S230: For each of the at least one structured field, obtain a structured processing result of the text to be processed corresponding to that structured field, whereby the structured processing result indicates the answer among the plurality of answers that corresponds to the question-and-answer statement in the text to be processed for the question corresponding to that structured field; and

[0045] Step S240: Based on the structured processing results of the text to be processed corresponding to each of the at least one structured field, obtain the structured information of the text to be processed.

[0046] By obtaining at least one structured field corresponding to the question-and-answer statements in the text to be processed, each structured field corresponding to a question with multiple answers, and by obtaining the structured processing result of the text to be processed corresponding to one of the multiple answers for each question with a structured field, and obtaining the structured information of the text to be processed based on the multiple structured processing results of the text to be processed corresponding to multiple structured fields, the process of obtaining structured information is transformed into the process of obtaining answers to the questions corresponding to the structured fields in the text to be processed. In this process, the text to be processed is obtained by multi-classifying the multiple answers. This makes the method of extracting structured information from the text to be processed simple. At the same time, since the results obtained by multi-classifying the text are accurate, the obtained structured information is accurate.

[0047] In related technologies, structural information in text is extracted based on structured fields that simultaneously include questions and corresponding answers. These structured fields are processed into classification labels (0 or 1) for model training. The trained model learns the implicit relationships between the input dialogue and the labeled tags, and can obtain the corresponding structured fields based on the input dialogue, using these structured fields as the structured information of the input dialogue. For example, the question "Do you drink alcohol?" and its corresponding answer "Drink" or "Don't drink" are processed into the structured fields "[Do you drink alcohol, Drink]" and "[Do you drink alcohol, Don't drink]", respectively. These two structured fields are then processed into classification labels to classify the text, thereby obtaining the corresponding structured fields of the text.

[0048] In the methods described above, adding new structured fields requires retraining the model. As the number of structured fields increases, the number of labels also increases, leading to a greater number of classifications, which hinders the expansion of structured fields. Furthermore, because structured fields are diverse and may be related (e.g., hyponyms, synonyms, etc.), simply treating them as numerical labels ignores the inherent textual information within the structured fields, treating them as independent labels for model training and prediction, thus losing semantic information.

[0049] According to the embodiments of this disclosure, only the question is converted into a structured field, and the multiple answers corresponding to the question are converted into category labels. The text is then processed to classify the text for multiple answers. This fully utilizes the semantic information of the text and the structured field. Even for newly added structured fields, the semantic information of the newly added structured fields can be used to classify the text and obtain accurate structured processing results. This avoids the problem of an increase in the number of categories caused by the increase in the number of structured fields and is conducive to the expansion of structured fields.

[0050] For example, the question "Do you drink alcohol?" is processed into a structured field "[Do you drink alcohol?]", with corresponding answers: "Yes", "No", and "Unknown". After obtaining the text to be processed, in response to the determination that the text contains question-and-answer statements corresponding to the structured field "[Do you drink alcohol?]", it is categorized among the three answers "Yes", "No", and "Unknown" to obtain the answer to the question "Do you drink alcohol?", thereby extracting the structural information in the text. In this process, the answers "Yes", "No", and "Unknown" are processed as labels. By utilizing the semantic information of the text to be processed and the structured field "[Do you drink alcohol?]", the text is categorized, improving the accuracy of the obtained structured information. Furthermore, when new structured fields are obtained, the semantic information of the new structured fields and the categorization process of processing the answers "Yes", "No", and "Unknown" as labels can still be utilized, avoiding the problem of an increased number of categories due to the increase in the number of structured fields.

[0051] In some embodiments, the text to be processed can be text containing question-and-answer statements between two objects, obtained in any manner. The objects are any objects capable of dialogue, such as a person or a voice output device.

[0052] In some embodiments, the text to be processed may be news text, which includes the dialogue text between a journalist and an interviewee.

[0053] In some embodiments, the text processing method according to this disclosure is applied to a medical scenario, and the text to be processed can be the dialogue text between a doctor and a patient. In one example, the text to be processed is obtained by converting the voice follow-up voice between an intelligent voice customer service representative and a patient into voice-to-text.

[0054] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0055] In some embodiments, the structured field set is a dataset pre-collected based on the use case of the text processing method. For example, if the use case of the text processing method is a medical scenario, then the structured field set is a dataset collected for the information that doctors need to know in the medical scenario, including structured fields derived from commonly asked questions by doctors in the medical scenario.

[0056] In embodiments according to this disclosure, each structured field in the structured field set corresponds to a question, and the question corresponding to each structured field corresponds to the same multiple answers.

[0057] For example, each structured field in the structured field set is transformed from a "yes or no..." question. Therefore, the answer to the question corresponding to each structured field includes three cases: "yes", "no", and "unknown".

[0058] In one example, the text processing method is used in a medical setting, with structured fields including fields such as "whether you drink alcohol" and "whether you take vitamins".

[0059] In some embodiments, step S220, obtaining at least one structured field corresponding to the text to be processed from the structured field set, includes: extracting keywords from the text to be processed; and determining the structured field as the structured field corresponding to the text to be processed in response to determining that a structured field in the structured field set contains the keyword.

[0060] In some other embodiments, step S220, obtaining at least one structured field corresponding to the text to be processed from the structured field set, includes: obtaining the similarity between the text to be processed and each structured field in the structured field set; in response to determining that the similarity between a structured field in the structured field set and the text to be processed is greater than a preset value, determining the structured field as the structured field corresponding to the text to be processed.

[0061] In some embodiments, in step S230, the text to be processed is classified for multiple answers to obtain the corresponding structured processing result of the text to be processed.

[0062] For example, in response to determining that the text to be processed contains words corresponding to the answer "yes", such as "yes" or "have", the structured processing result of the text to be processed indicates that the text to be processed corresponds to the answer "yes".

[0063] In some embodiments, such as Figure 3 As shown, step S230, obtaining the structured processing result of the text to be processed corresponding to each of the at least one structured field, includes:

[0064] Step S310: Obtain the feature vector corresponding to the text to be processed;

[0065] Step S320: Obtain the feature vector corresponding to each of the at least one structured field;

[0066] Step S330: For each of the at least one structured fields, based on the feature vector corresponding to the structured field and the feature vector corresponding to the text to be processed, obtain the structured processing result of the text to be processed corresponding to the structured field.

[0067] By obtaining the feature vectors of each structured field and the text to be processed, and based on the feature vectors of each structured field and the text to be processed, obtaining the question corresponding to that structured field and the corresponding answer of the text to be processed among multiple answers, it makes full use of the semantic information obtained in the text processing process and improves the accuracy of the obtained structured processing results.

[0068] In some embodiments, in step S310, a trained feature vector extraction network is used to obtain the feature vector corresponding to the text to be processed. This feature vector extraction network can be any natural language processing model.

[0069] In some embodiments, obtaining the feature vector corresponding to the text to be processed in step S310 includes:

[0070] By inputting the text to be processed into a first vector processing network, the feature vector corresponding to the text to be processed is obtained; wherein,

[0071] The first vector processing network is obtained by training a dual-tower model containing the first vector processing network, the second vector processing network, and a classification network using training text and at least one structured field corresponding to the training text.

[0072] Each of the plurality of training texts includes a question-and-answer statement. Each of the at least one structured field corresponds to at least one question in the question-and-answer statement of the training text. The first vector processing network is used to obtain the feature vector corresponding to the training text. The second vector network is used to obtain the feature vector corresponding to each of the at least one structured field. The classification network is used to obtain a corresponding classification result based on the feature vector corresponding to each of the at least one structured field and the feature vector corresponding to the corresponding training text. The classification result indicates the answer in the plurality of answers that corresponds to the question-and-answer statement of the training text for the question corresponding to the structured field.

[0073] The trained dual-tower model obtains feature vectors corresponding to the text to be processed. During training, the dual-tower model inputs the training text into a first vector processing network and then into a second vector processing network via structured fields to obtain feature vectors corresponding to the training text and structured fields respectively. Simultaneously, the feature vectors obtained from the first and second vector processing networks are input into a classification network to obtain the answer corresponding to the question in the structured field processed by the second vector network. This makes the training process of the dual-tower model a classification process for the training text and the corresponding structured fields. As a result, the trained dual-tower model obtains feature vectors that can accurately obtain classification results based on the feature vectors of the input text and structured fields, improving the accuracy of the obtained feature vectors and thus improving the accuracy of obtaining the structured processing results corresponding to the text to be processed based on the feature vectors.

[0074] See Figure 4 The diagram illustrates a schematic of a dual-tower model involved in a text processing method according to some embodiments of the present disclosure. The dual-tower model 400 includes a first vector processing network 410 and a second vector processing network 420. The first vector processing network 410 processes the input text "Doctor: How has your sleep been lately? Patient: I slept quite well" to obtain a corresponding feature vector u. The second vector processing network 420 processes the input structured field "Whether you slept well" to obtain a corresponding feature vector v. The dual-tower model 400 also includes a classification network 430 for classification based on the feature vectors u and v to obtain a classification result indicating that the input text corresponds to one of "yes," "no," or "unknown."

[0075] In some embodiments, both the first vector processing network and the second vector processing network include an ERNIE network and a pooling network. In some embodiments, the classification network includes a softmax network.

[0076] In some embodiments, after obtaining the similarity vector between feature vector u and feature vector v, the feature vector u, feature vector v, and the similarity vector between them are input into a softmax network to obtain the classification result.

[0077] In some embodiments, in step S320, each structured input is fed into the second vector processing network to obtain the feature vector corresponding to the structured field, and in step S330, for each structured field, the classification network is used to obtain the structured processing result of the text to be processed corresponding to the structured field based on the feature vector corresponding to the structured field and the feature vector corresponding to the text to be processed.

[0078] In some embodiments, after the dual-tower model is trained, each structured field in the structured field set is input into the second vector network to obtain the feature vector corresponding to that structured field and store it in the database to obtain a feature vector set. Therefore, in step S320, the feature vector corresponding to each of the at least one structured field is obtained from the feature vector set including the feature vectors corresponding to the plurality of structured fields.

[0079] By pre-obtaining the feature vectors corresponding to each structured field in the structured field set, during the processing of the text to be processed, it is only necessary to retrieve the feature vectors corresponding to each structured field in at least one structured field of the text to be processed from the feature vector set, without recalculating the feature vectors corresponding to the structured fields. At the same time, only the first vector processing network and the classification network in the trained dual-tower model mentioned above need to be used for calculation to obtain the structured processing result of the text to be processed. This reduces the amount of computation required to calculate the feature vectors of the structured fields and improves the speed of obtaining the structured processing result of the text to be processed.

[0080] In some embodiments, the first structured field in the at least one structured field includes attribute words, such as... Figure 5 As shown, step S230, obtaining the structured processing result of the text to be processed corresponding to each of the at least one structured field, includes:

[0081] Step S510: In response to determining that the text to be processed includes a semantic association word corresponding to the attribute word, replace the semantic association word in the text to be processed with the attribute word to obtain updated text; and

[0082] Step S520: Based on the updated text, obtain the structured processing result of the text to be processed corresponding to each of the at least one structured field.

[0083] Considering that in real-world language environments, many different words often have certain relationships, causing them to point to the same semantic meaning, in order to improve the accuracy of structured processing, for such words with relationships, we statistically analyze the attribute words and the semantic related words (e.g., entity words indicating a certain entity) in the relationship using prior knowledge. During processing, we replace the semantic related words with the corresponding attribute words, so that the obtained structured processing result further utilizes the relationship between attribute words and semantic related words based on prior knowledge, thereby improving the accuracy of the obtained structured processing result.

[0084] For example, regarding the structured field "Whether or not you are taking vitamins", which includes the attribute word "vitamin", when the text to be processed is "Doctor: What vitamins are you taking specifically? Patient: I'm taking Centrum recently", the "Centrum" and "vitamin" are not directly related. In the embodiments according to this disclosure, based on prior knowledge, it can be determined that there is a semantic association between "Centrum" and "vitamin", so that "Centrum" in the text to be processed can be replaced with "vitamin" for processing (e.g., processing is performed using steps S310-S330) to obtain the corresponding structured processing result.

[0085] After steps S210-S230, the processed text obtains a structured processing result corresponding to each of at least one structured field. The structured processing result can be, for example, a combination of the structured field and the answer indicated by that result. For instance, the structured processing result could be "Is sleep poor? Yes".

[0086] In step S240, the structured processing results of the text to be processed corresponding to each of the at least one structured field are provided to the user as structured information.

[0087] In some embodiments, see Figure 6 Step S240, obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field includes:

[0088] Step S610: Obtain the association relationship between the first structured field and the second structured field in the at least one structured field, the association relationship including causal relationship or parallel relationship;

[0089] Step S620: Based on the association relationship, merge the structured processing results corresponding to the first structured field and the structured processing results corresponding to the second structured field to obtain a merged result; and

[0090] Step S630: Based on the merging result, obtain the structured information.

[0091] Considering that structured fields have hierarchical and parallel relationships, merging the structured processing corresponding to each structured field during the process of obtaining structured information, and obtaining structured information based on the merged result, can simplify the expression of structured information.

[0092] For example, the text to be processed corresponds to two structured fields: "Poor sleep" and "Excessive sleep," which have a causal relationship. When the structured processing results for the two result fields are "Poor sleep is yes" and "Excessive sleep is yes," the two structured processing results can be merged into "Poor sleep is yes: Excessive sleep" as structured information, making the expression of structured information clear and simplified.

[0093] For example, the text to be processed corresponds to two structured fields: "Is diastolic blood pressure higher than 80?" and "Is systolic blood pressure higher than 130?", which are parallel fields. When the structured processing results for the two result fields are "Is diastolic blood pressure higher than 80?" and "Is systolic blood pressure higher than 130?", these results are combined into "Diastolic blood pressure higher than 80 / Systolic blood pressure higher than 130?" as the final result information, making the expression of structured information clear and simplified.

[0094] In some embodiments, the text processing method according to this disclosure further includes: in response to a numeric word in the text to be processed that includes an indicative value, performing structured processing on the numeric word to obtain a structured processing result corresponding to the numeric word, wherein, in step S240, during the process of obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field, the structured information of the text to be processed is also obtained based on the structured processing result corresponding to the numeric word.

[0095] In some embodiments, the conversion from Chinese to Arabic numerals is achieved using a tool that converts between Chinese and Arabic data. For example, by iterating in reverse order from right to left, the numerical values ​​are accumulated by multiplying the numbers by the units to achieve the conversion from Chinese to Arabic numerals.

[0096] Given that question-and-answer statements in texts are often presented in a conversational style between the communicators, words indicating numerical values ​​often include vague numerical ranges. In some embodiments, such as Figure 7 As shown, the text processing method according to this disclosure also includes:

[0097] Step S710: Determine whether the text to be processed includes numerical range words corresponding to the numerical range;

[0098] Step S720: In response to determining that the text to be processed includes numerical range words corresponding to the numerical range, a second structured processing result of the text to be processed is obtained, the second structured processing result including the numbers corresponding to the numerical range, and wherein, step S240: obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field includes:

[0099] The structured information of the text to be processed is obtained based on the structured processing result of each of the at least one structured field and the second structured processing result.

[0100] By performing structured processing on numerical range words corresponding to numerical ranges, it is possible to perform structured processing on fuzzy numerical ranges during the extraction of structured information from the text to be processed, and obtain the corresponding structured processing results. In the process of obtaining structured information, further structured information is obtained based on the structured processing results obtained from the fuzzy numerical range, thereby further improving the accuracy of the obtained structured information.

[0101] In some embodiments, in response to determining that multiple characters indicating numerical values ​​appear in a preset number of consecutive characters, such that the language expression conforms to strict numerical specifications, the preset number of consecutive characters are determined as numerical range words.

[0102] For example, when the text to be processed contains characters such as "thirty-forty", "two or three", "twenty-three or four", and "three to four", "thirty-forty", "two or three", "twenty-three or four", and "three to four" are identified as numerical range words that indicate the numerical range.

[0103] Considering that in some cases, such as numerical expressions with units in the middle or decimal places, for example, "height is 1.67 meters" or "daily water intake is 1.34 liters," which are accurate values, according to some embodiments of this disclosure, in response to the fact that the numerical range term includes at least one unit of measurement from a plurality of units of measurement, it is determined that the text to be processed does not include numerical range terms corresponding to the numerical range. Therefore, it is structured according to the method for structuring numerical terms described above, in order to reduce the number of structuring steps and the amount of data processed.

[0104] In some embodiments, after structuring the numerical range terms, the resulting second structured result is a standardized numerical range represented by Arabic numerals. For example, after structuring "thirty-forty", the resulting second structured result is "30-40".

[0105] In some embodiments, the text processing method according to this disclosure further includes: in response to a time word indicating time in the text to be processed, performing structured processing on the time word to obtain a structured processing result corresponding to the time word, wherein, in step S240, during the process of obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field, the structured information of the text to be processed is also obtained based on the structured processing result corresponding to the time word.

[0106] For example, the time parsing module can be used to parse time words indicating year, month, day, week, season, quarter, holiday, etc., to obtain the corresponding structured processing results.

[0107] Given that question-and-answer statements in text are often presented in a conversational style between the communicators, time indicators often include time range indicators that specify a vague time range, such as "the last week," "this month," etc., in some embodiments according to this disclosure, the text processing method further includes:

[0108] In response to determining that the text to be processed includes time range words corresponding to a time range, a third structured processing result of the text to be processed is obtained, the third structured processing result indicating the time range, wherein in step S240, obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field includes:

[0109] The structured information of the text to be processed is obtained based on the structured processing result of each of the at least one structured field and the third structured processing result.

[0110] By performing structured processing on time range words corresponding to the time range, it is possible to perform structured processing on fuzzy time ranges during the extraction of structured information from the text to be processed, and obtain the corresponding structured processing results. In the process of obtaining structured information, further structured information is obtained based on the structured processing results obtained from the fuzzy time range, thereby further improving the accuracy of the obtained structured information.

[0111] In some embodiments, in response to the inclusion of trigger words and time entity words in the text to be processed, it is determined that the text to be processed includes time range words corresponding to the time range.

[0112] In some embodiments, the trigger word may be a word that indicates an approximate or vague concept, such as "recently", "the last few", "the previous one", etc.

[0113] In some embodiments, time entity words such as “*hour”, “day”, “week”, “month”, “quarter”, etc., can be used as words for time measurement units.

[0114] In some embodiments, a third structured processing result is obtained by combining trigger words and time entity words in the text to be processed.

[0115] For example, for the text to be processed, "Doctor: How was your sleep? Patient: I haven't been sleeping well in the last week", the time entity word "week" and the trigger word "recently" are obtained. By combining the two, the third structured processing result, "[last week]", is obtained.

[0116] In some embodiments, structured information is obtained by merging a second structured processing result obtained by structuring numerical words, a third structured processing result obtained by structuring time words, and the structured processing result of the text to be processed corresponding to each of the at least one structured field to obtain structured information, which is then displayed to the user.

[0117] According to another aspect of this disclosure, a text processing apparatus is also provided, see [link to relevant documentation]. Figure 8 The apparatus 800 includes: a text-to-process acquisition unit 810 configured to acquire text to be processed, the text including a question-and-answer statement between two objects; a structured field acquisition unit 820 configured to acquire at least one structured field corresponding to the text to be processed from a set of structured fields, each structured field in the set corresponding to a question and a plurality of answers; a structured processing result acquisition unit 830 configured to acquire a structured processing result of the text to be processed corresponding to each of the at least one structured field, the structured processing result indicating the answer corresponding to the question-and-answer statement of the text to be processed among the plurality of answers for the question corresponding to the structured field; and a structured information acquisition unit 840 configured to acquire structured information of the text to be processed based on the structured processing result of the text to be processed corresponding to each of the at least one structured field.

[0118] In some embodiments, the structured processing result acquisition unit includes: a first feature vector acquisition unit configured to acquire a feature vector corresponding to the text to be processed; a second feature vector acquisition unit configured to acquire a feature vector corresponding to each of the at least one structured field; and a structured processing result acquisition subunit configured to, for each of the at least one structured field, acquire a structured processing result corresponding to that structured field based on the feature vector corresponding to that structured field and the feature vector corresponding to the text to be processed.

[0119] In some embodiments, the first feature vector acquisition unit includes: a text input unit configured to obtain a feature vector corresponding to the text to be processed by inputting the text to be processed into a first vector processing network; wherein the first vector processing network is obtained by training a dual-tower model including the first vector processing network, a second vector processing network, and a classification network using training text and at least one structured field corresponding to the training text; wherein each of the plurality of training texts includes a question-and-answer statement, each of the at least one structured field corresponds to at least one question in the question-and-answer statement in the training text; the first vector processing network is used to obtain the feature vector corresponding to the training text; the second vector network is used to obtain the feature vector corresponding to each of the at least one structured field; and the classification network is used to obtain a corresponding classification result based on the feature vector corresponding to each of the at least one structured field and the feature vector corresponding to the corresponding training text, the classification result indicating the answer corresponding to the question-and-answer statement in the training text among the plurality of answers for the question corresponding to the structured field.

[0120] In some embodiments, the second feature vector acquisition unit includes: a retrieval unit configured to obtain a feature vector corresponding to each of the at least one structured field from a feature vector set including feature vectors corresponding to the plurality of structured fields, wherein each feature vector in the feature vector set is obtained by inputting the corresponding structured field into the second vector processing network.

[0121] In some embodiments, the first structured field in the at least one structured field includes an attribute word, and the structured processing result acquisition unit includes: a first response unit configured to replace the semantic association word in the text to be processed with the attribute word in response to determining that the text to be processed includes a semantic association word corresponding to the attribute word, so as to obtain an updated text; and a first acquisition subunit configured to obtain a structured processing result of the text to be processed corresponding to each structured field in the at least one structured field based on the updated text.

[0122] In some embodiments, the structured information acquisition unit includes: an association relationship acquisition unit configured to obtain an association relationship between a first structured field and a second structured field in the at least one structured field, the association relationship including a causal relationship or a parallel relationship; a merging unit configured to merge the structured processing results corresponding to the first structured field and the second structured field based on the association relationship to obtain a merged result; and a second acquisition subunit configured to obtain the structured information based on the merged result.

[0123] In some embodiments, the apparatus 800 further includes: a first determining unit configured to determine whether the text to be processed includes a numerical range word corresponding to a numerical range; a second responding unit configured to obtain a second structured processing result of the text to be processed in response to determining that the text to be processed includes a numerical range word corresponding to a numerical range, the second structured processing result including a number corresponding to the numerical range; and wherein the structured information acquisition unit includes: a third acquisition subunit configured to obtain structured information of the text to be processed based on the structured processing result of each of the at least one structured field and the second structured processing result.

[0124] In some embodiments, the first determining unit includes a third response unit configured to determine, in response to the numeric range word including at least one of a plurality of units of measurement, that the text to be processed does not include numeric range words corresponding to the numeric range.

[0125] In some embodiments, the apparatus 800 further includes: a fourth response unit configured to obtain a third structured processing result of the text to be processed in response to determining that the text to be processed includes time range words corresponding to a time range, the third structured processing result indicating the time range, and wherein the structured information acquisition unit includes: a fourth acquisition subunit configured to obtain structured information of the text to be processed based on the structured processing result of each of the at least one structured field and the third structured processing result.

[0126] According to embodiments of this disclosure, an electronic device, a readable storage medium, and a computer program product are also provided.

[0127] refer to Figure 9The present invention describes a structural block diagram of an electronic device 900 that can serve as a server or client of the present disclosure, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0128] like Figure 9 As shown, the electronic device 900 includes a computing unit 901, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 902 or a computer program loaded into a random access memory (RAM) 903 from a storage unit 908. The RAM 903 may also store various programs and data required for the operation of the electronic device 900. The computing unit 901, ROM 902, and RAM 903 are interconnected via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0129] Multiple components in electronic device 900 are connected to I / O interface 905, including: input unit 906, output unit 907, storage unit 908, and communication unit 909. Input unit 906 can be any type of device capable of inputting information to electronic device 900. Input unit 906 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of electronic device, and can include, but is not limited to, a mouse, keyboard, touchscreen, trackpad, trackball, joystick, microphone, and / or remote control. Output unit 907 can be any type of device capable of presenting information, and can include, but is not limited to, a monitor, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 908 can include, but is not limited to, disk and optical disk. Communication unit 909 allows electronic device 900 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and can include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers and / or chipsets, such as Bluetooth™ devices, 802.11 devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0130] The computing unit 901 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 901 performs the various methods and processes described above, such as method 200. For example, in some embodiments, method 200 may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 908. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 900 via ROM 902 and / or communication unit 909. When the computer program is loaded into RAM 903 and executed by the computing unit 901, one or more steps of method 200 described above may be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform method 200 by any other suitable means (e.g., by means of firmware).

[0131] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0132] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0133] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0134] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0135] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0136] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0137] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0138] While embodiments or examples of this disclosure have been described with reference to the accompanying drawings, it should be understood that the methods, systems, and devices described above are merely exemplary embodiments or examples, and the scope of the invention is not limited by these embodiments or examples, but only by the granted claims and their equivalents. Various elements in the embodiments or examples may be omitted or replaced by their equivalents. Furthermore, the steps may be performed in a different order than that described in this disclosure. Further, various elements in the embodiments or examples may be combined in various ways. Importantly, as the technology evolves, many elements described herein can be replaced by equivalents that appear after this disclosure.

Claims

1. A text processing method, comprising: Obtain the text to be processed, which includes a question-and-answer statement between two objects; From the set of structured fields, obtain at least one structured field corresponding to the text to be processed, where each structured field in the set of structured fields corresponds to a question, and the question corresponds to multiple pre-defined answers as category labels; For each of the at least one structured field, a structured processing result corresponding to that structured field is obtained from the text to be processed. This structured processing result indicates the answer among a plurality of answers to the question and answer statement of the text to be processed, for the question corresponding to that structured field. The step of obtaining the structured processing result corresponding to each of the at least one structured field includes: By inputting the text to be processed into a first vector processing network, the feature vector corresponding to the text to be processed is obtained; Obtain the feature vector corresponding to each of the at least one structured field, wherein the feature vector corresponding to the structured field is obtained by inputting the structured field into a second vector processing network; and For each of the at least one structured field, the feature vector corresponding to that structured field and the feature vector corresponding to the text to be processed are input into a classification network. A classification result is obtained based on the classification network. The classification result indicates the answer to that structured field among multiple answers corresponding to that structured field, and is used as the structured processing result of that structured field. Based on the structured processing results of the text to be processed corresponding to each of the at least one structured field, the structured information of the text to be processed is obtained.

2. The method according to claim 1, wherein, The first vector processing network is obtained by training a dual-tower model containing the first vector processing network, the second vector processing network, and a classification network using training text and at least one structured field corresponding to the training text. Each of the plurality of training texts includes a question-and-answer statement. Each of the at least one structured field corresponds to at least one question in the question-and-answer statement of the training text. The first vector processing network is used to obtain the feature vector corresponding to the training text. The second vector processing network is used to obtain the feature vector corresponding to each of the at least one structured field. The classification network is used to obtain a corresponding classification result based on the feature vector corresponding to each of the at least one structured field and the feature vector corresponding to the corresponding training text. The classification result indicates the answer in the plurality of answers that corresponds to the question-and-answer statement of the training text for the question corresponding to the structured field.

3. The method according to claim 2, wherein, Obtaining the feature vector corresponding to each of the at least one structured field includes: From a feature vector set comprising multiple feature vectors corresponding to multiple structured fields, a feature vector corresponding to each of the at least one structured field is obtained, wherein each feature vector in the feature vector set is obtained by inputting the corresponding structured field into the second vector processing network.

4. The method according to claim 1, wherein, The first structured field in the at least one structured field includes attribute words, and obtaining the structured processing result of the text to be processed corresponding to each structured field in the at least one structured field includes: In response to determining that the text to be processed includes a semantic association word corresponding to the attribute word, the semantic association word in the text to be processed is replaced with the attribute word to obtain updated text; and Based on the updated text, the structured processing result of the text to be processed corresponding to each of the at least one structured field is obtained.

5. The method according to claim 1, wherein obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field includes: Obtain the association relationship between the first structured field and the second structured field in the at least one structured field, wherein the association relationship includes a causal relationship or a parallel relationship; Based on the aforementioned relationship, the structured processing results corresponding to the first structured field and the structured processing results corresponding to the second structured field are merged to obtain a merged result; as well as Based on the merging result, the structured information is obtained.

6. The method according to any one of claims 1-5, further comprising: Determine whether the text to be processed includes numerical range words that correspond to the numerical range; In response to determining that the text to be processed includes numerical range words corresponding to the numerical range, a second structured processing result of the text to be processed is obtained, wherein the second structured processing result includes the numbers corresponding to the numerical range, and wherein, The process of obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field includes: The structured information of the text to be processed is obtained based on the structured processing result of each of the at least one structured field and the second structured processing result.

7. The method according to claim 6, wherein, Determining whether the text to be processed includes numeric range words corresponding to the numeric range includes: In response to the fact that the numerical range term includes at least one of a plurality of units of measurement, it is determined that the text to be processed does not include numerical range terms corresponding to the numerical range.

8. The method according to any one of claims 1-5, further comprising: In response to determining that the text to be processed includes time range words corresponding to a time range, a third structured processing result of the text to be processed is obtained, the third structured processing result indicating the time range, and wherein, The process of obtaining the structured information of the text to be processed based on the structured processing result of each of the at least one structured field includes: The structured information of the text to be processed is obtained based on the structured processing result of each of the at least one structured field and the third structured processing result.

9. A text processing apparatus, comprising: The text acquisition unit is configured to acquire text to be processed, which includes a question-and-answer statement between two objects. The structured field acquisition unit is configured to obtain at least one structured field corresponding to the text to be processed from a set of structured fields, wherein each structured field in the set of structured fields corresponds to a question, and the question corresponds to multiple preset answers as category labels. A structured processing result acquisition unit is configured to obtain, for each of the at least one structured field, a structured processing result corresponding to that structured field in the text to be processed. The structured processing result indicates the answer corresponding to the question and answer statement of the text to be processed among a plurality of answers for the question corresponding to that structured field. The structured processing result acquisition unit includes: The first feature vector acquisition unit is configured to obtain the feature vector corresponding to the text to be processed by inputting the text to be processed into the first vector processing network; The second feature vector acquisition unit is configured to obtain a feature vector corresponding to each of the at least one structured field, wherein the feature vector corresponding to the structured field is obtained by inputting the structured field into a second vector processing network; and A structured processing result acquisition subunit is configured to, for each of the at least one structured field, input the feature vector corresponding to that structured field and the feature vector corresponding to the text to be processed into a classification network, obtain a classification result based on the classification network, the classification result indicating the answer to that structured field among multiple answers corresponding to that structured field, and serve as the structured processing result for that structured field; and The structured information acquisition unit is configured to obtain the structured information of the text to be processed based on the structured processing result corresponding to each of the at least one structured field in the text to be processed.

10. The apparatus according to claim 9, wherein, The first vector processing network is obtained by training a dual-tower model containing the first vector processing network, the second vector processing network, and a classification network using training text and at least one structured field corresponding to the training text. Each of the plurality of training texts includes a question-and-answer statement. Each of the at least one structured field corresponds to at least one question in the question-and-answer statement of the training text. The first vector processing network is used to obtain the feature vector corresponding to the training text. The second vector processing network is used to obtain the feature vector corresponding to each of the at least one structured field. The classification network is used to obtain a corresponding classification result based on the feature vector corresponding to each of the at least one structured field and the feature vector corresponding to the corresponding training text. The classification result indicates the answer in the plurality of answers that corresponds to the question-and-answer statement of the training text for the question corresponding to the structured field.

11. The apparatus according to claim 10, wherein, The second feature vector acquisition unit includes: The retrieval unit is configured to obtain a feature vector corresponding to each of the at least one structured field from a feature vector set comprising multiple feature vectors corresponding to multiple structured fields, wherein each feature vector in the feature vector set is obtained by inputting the corresponding structured field into the second vector processing network.

12. The apparatus according to claim 9, wherein, The first structured field in the at least one structured field includes attribute words, and the structured processing result acquisition unit includes: A first response unit is configured to, in response to determining that the text to be processed includes a semantic association word corresponding to the attribute word, replace the semantic association word in the text to be processed with the attribute word to obtain an updated text; and The first acquisition subunit is configured to obtain, based on the updated text, the structured processing result of the text to be processed corresponding to each of the at least one structured field.

13. The apparatus according to claim 9, wherein, The structured information acquisition unit includes: The association relationship acquisition unit is configured to obtain the association relationship between a first structured field and a second structured field in the at least one structured field, wherein the association relationship includes a causal relationship or a parallel relationship; A merging unit is configured to merge the structured processing results corresponding to the first structured field and the structured processing results corresponding to the second structured field based on the association relationship, to obtain a merged result; and The second acquisition subunit is configured to obtain the structured information based on the merging result.

14. The apparatus according to any one of claims 9-13, further comprising: The first determining unit is configured to determine whether the text to be processed includes a numerical range word corresponding to the numerical range. The second response unit is configured to, in response to determining that the text to be processed includes numeric range words corresponding to a numeric range, obtain a second structured processing result of the text to be processed, wherein the second structured processing result includes numbers corresponding to the numeric range, and wherein... The structured information acquisition unit includes: The third acquisition subunit is configured to obtain the structured information of the text to be processed based on the structured processing result of each of the at least one structured field and the second structured processing result.

15. The apparatus according to claim 14, wherein, The first determining unit includes: The third response unit is configured to determine, in response to the fact that the numerical range word includes at least one of a plurality of units of measurement, that the text to be processed does not include numerical range words corresponding to the numerical range.

16. The apparatus according to any one of claims 9-13, further comprising: A fourth response unit is configured to, in response to determining that the text to be processed includes time range words corresponding to a time range, obtain a third structured processing result of the text to be processed, the third structured processing result indicating the time range, and wherein, The structured information acquisition unit includes: The fourth acquisition subunit is configured to obtain the structured information of the text to be processed based on the structured processing result of each of the at least one structured field and the third structured processing result.

17. An electronic device comprising: At least one processor; as well as A memory that is communicatively connected to the at least one processor; in The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

18. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-8.

19. A computer program product comprising a computer program, wherein, When the computer program is executed by a processor, it implements the method of any one of claims 1-8.

Images