CN111858725A - Method and system for determining event attributes - Google Patents

Abstract

The embodiments of the present application disclose a method, system, device and storage medium for determining an event attribute. The method includes: acquiring work order information generated according to an event description of a target event, the work order information including at least structured information and narrative information; obtaining a first feature according to the narrative information; obtaining a second feature; fusing the first feature and the second feature to obtain a fused feature; and obtaining an event attribute of the target event based on the fused feature. The present application can improve the accuracy of event attribute determination.

Description

Method and system for determining event attributes

technical field

The present application relates to the technical field of data processing, and in particular, to a method and system for determining an event attribute.

Background technique

With the development of the shared transportation service industry, people are more and more accustomed to handling security issues through customer service, but the ability of customer service to handle security incidents is limited. In addition, subjective factors have a great influence on the judgment results of customer service, which will lead to misjudgment of security incidents. Therefore, it is necessary to provide an event attribute determination method and system.

SUMMARY OF THE INVENTION

A first aspect of the present application provides an event attribute determination method. The method for detecting and determining an event attribute includes: acquiring work order information generated according to an event description of a target event, where the work order information at least includes structured information and narrative information; obtaining a first feature according to the narrative information; The structured information is used to obtain a second feature; the first feature and the second feature are fused to obtain a fusion feature; and an event attribute of the target event is obtained based on the fusion feature.

A second aspect of the present application provides an event attribute determination system, including: a work order acquisition module, configured to acquire work order information generated according to an event description of a target event, where the work order information at least includes structured information and narrative information The first feature extraction module is used to obtain the first feature according to the narrative information; the second feature extraction module is used to obtain the second feature according to the structured information; the feature fusion module is used to fuse the first feature The feature and the second feature are used to obtain a fusion feature; and a determination module is used to obtain the event attribute of the target event based on the fusion feature.

In some embodiments, the ticket information is generated by the customer service based on the event description and/or the processing result of the event.

In some embodiments, the structured information in the work order information includes at least one of the following information: order number, license plate number, phone number, whether to call the police, whether the police file a case, whether the user requests processing, and The degree of urgency required for processing; the narrative information in the work order information includes at least one of the following information: a user's description of the event, a description of the police's processing result, and a description of the customer service's processing result.

In some embodiments, the first feature extraction module uses a text transformation model to process the narrative information to obtain the first feature.

In some embodiments, the text transformation model includes at least one of the following deep learning models: Fasttext, HAN, Text CNN, Transformer, LR, and XG Boost.

In some embodiments, a training module is further included, and the training module is used to obtain a sample set, the sample set includes a plurality of pieces of work order information, and the descriptive information in the work order information is used as input, and given by customer service The event attribute corresponding to the work order information is used as an identifier to train the text conversion model.

In some embodiments, the second feature extraction module generates the second feature by processing structured information through an extraction model.

In some embodiments, the extraction model employs a rule-based and/or AC automaton-based approach to feature extraction.

In some embodiments, the feature fusion module directly splices the first feature and the second feature or processes the first feature and the second feature through a set algorithm to generate a combined feature to obtain a fused feature.

In some embodiments, the determining module processes the fusion feature using a classification model to obtain a classification result of the event attribute; the classification result of the event attribute includes at least one of the following: whether the event is safe, whether the event is accurate, whether the event is accurate Traceability, whether events are repeatable.

In some embodiments, the classification model includes at least one of the following deep learning models: XG Boost, GBDT, Adaboost, Random Forest.

In some embodiments, a training module is further included, and the training module is used to obtain a sample set, the sample set includes a plurality of pieces of work order information, and the fusion feature corresponding to the work order information is used as an input, and the information provided by the customer service is used as the input. The event attribute corresponding to the work order information is used as an identifier to train the classification model.

A third aspect of the present application provides an event attribute determination device, comprising at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute at least one of the computer instructions part of the instructions to implement the operations described above.

A fourth aspect of the present application provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the operations described above.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following briefly introduces the accompanying drawings that are used in the description of the embodiments. Obviously, the accompanying drawings in the following description are only some examples or embodiments of the present application. For those of ordinary skill in the art, without any creative effort, the present application can also be applied to the present application according to these drawings. other similar situations. in:

1 is a schematic diagram of an application scenario of an event attribute determination system according to some embodiments of the present application;

FIG. 2 is a schematic block diagram of an event attribute determination system according to some embodiments of the present application;

FIG. 3 is an exemplary flowchart of a method for determining an event attribute according to some embodiments of the present application; and

FIG. 4 is an example of common structured information and narrative information according to some embodiments of the present application.

Detailed ways

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following briefly introduces the accompanying drawings that are used in the description of the embodiments. Obviously, the accompanying drawings in the following description are only some examples or embodiments of the present application. For those of ordinary skill in the art, without any creative effort, the present application can also be applied to the present application according to these drawings. other similar situations. Unless obvious from the locale or otherwise specified, the same reference numbers in the figures represent the same structure or operation.

As shown in this application and in the claims, unless the context clearly dictates otherwise, the words "a", "an", "an" and/or "the" are not intended to be specific in the singular and may include the plural. Generally speaking, the terms "comprising" and "comprising" only imply that the clearly identified steps and elements are included, and these steps and elements do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

While this application makes various references to certain modules in systems according to embodiments of the application, any number of different modules may be used and run on a vehicle client and/or server. The modules are illustrative only, and different aspects of the systems and methods may use different modules.

Flow diagrams are used in this application to illustrate operations performed by a system according to an embodiment of the application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Rather, the various steps may be processed in reverse order or concurrently. At the same time, other actions can be added to these procedures, or a step or steps can be removed from these procedures.

Furthermore, although the systems and methods disclosed in this application are primarily described with respect to transportation services, it should be understood that this is merely an exemplary embodiment. The system and method of the present invention may be applicable to any other on-demand services, such as housekeeping services, takeaway services, and the like. In some embodiments, the systems and methods of the present application may be applied to different transportation systems, including land, marine, aerospace, etc., or any combination of the foregoing examples. Vehicles used in the transportation system may include taxis, private cars, rideshares, buses, trains, bullet trains, high-speed rail, subways, ships, airplanes, spacecraft, hot air balloons, unmanned vehicles, etc., or combinations thereof. The transportation system may also include any transportation system suitable for management and/or distribution, eg, systems used to transmit and/or receive couriers. The application scenarios of the system or method of the present application may also include web pages, browser plug-ins, client terminals, customized systems, internal analysis systems, artificial intelligence robots, etc., or any combination thereof.

It is understandable that in some practical application scenarios, customer service needs to process a large number of events that occur during the on-demand service process every day, and quickly provide the processing results. Just as an example, in a shared transportation service, users rely on customer service to handle security incidents, but the actual ability of customer service to handle security incidents is limited. For example, a customer service may rate an incident with potential security risks as a normal incident. In some embodiments, the relevant information of the security events processed by the customer service can be processed through a computer algorithm to find out whether the security events processed by the customer service have missed security risks, and intervene to prevent the occurrence of a vicious car accident. . In some embodiments, the customer service process the security event based on the event description provided by the user to obtain work order information, the relevant event description and the processing result of the customer service can be included in the work order information corresponding to the event, and the work order is processed through a computer algorithm Information can be used to detect potential safety hazards again, thereby effectively reducing the work pressure of customer service.

FIG. 1 is a schematic diagram of an application scenario of an event attribute determination system according to some embodiments of the present application.

As shown in FIG. 1 , an exemplary event attribute determination system 100 may include a server 110 , a network 120 , a user terminal 130 and a storage module 140 .

In some embodiments, the server 110 may be used for event attribute determination. The server 110 may be a single server or a server group. A server farm can be centralized, such as a data center. A server farm can also be distributed, such as a distributed system. The server 110 may be local or remote. In some embodiments, server 110 may include a control processor 112 for executing instructions (program code) of server 110 . For example, the control processor 112 can execute the instructions of the event detection program, and then analyze and process the work order information through a certain algorithm to determine the detection result.

In some embodiments, terminal 130 includes a service requester terminal and/or a customer service terminal. For example only, the service requester may be the individual who initiates the car use request. In some embodiments, the service requester can send the event description of the target event that occurs during the transportation service to the server or the customer service terminal through its terminal. In some embodiments, the customer service terminal may receive the event description of the target event sent by the service requester terminal, process the target event, and generate work order information and/or processing results. In some embodiments, the work order information may include structured information generated based on the event description, narrative information, and/or event attributes initially judged by the customer. In some embodiments, the customer service terminal may return the processing result to the service requester terminal, or send the work order information to the server 110 for further processing. The terminal 130 includes, but is not limited to, one or a combination of a mobile phone 130-1, a tablet computer 130-2, a notebook computer 130-3, and the like. The server 110 can access the work order information stored in the storage module 140 , and can also transmit the detection result to the user terminal 130 through the network 120 .

In some embodiments, the storage module 140 may refer to a storage-capable device. The storage module 140 is mainly used to store event descriptions sent from the user terminal 130 and various data generated during the operation of the server 110 . The storage module 140 may be local or remote. The connection or communication between the system database and other modules of the system can also be wired or wireless. Network 120 may provide a channel for information exchange. The network 120 may be a single network or a combination of multiple networks. The network 120 may include, but is not limited to, one or a combination of a local area network, a wide area network, a public network, a private network, a wireless local area network, a virtual network, a metropolitan area network, a public switched telephone network, and the like. Network 120 may include a variety of network access points, such as wired or wireless access points, base stations (eg, 120-1, 120-2), or network switching points, through which data sources are connected to network 120 and through network 120 send Message.

Those of ordinary skill in the art will understand that when an element of the event attribute determination system 100 executes, the element may execute through electrical and/or electromagnetic signals. For example, when the user terminal 130 processes a task, such as when a user makes an event description, the user terminal 130 may operate logic circuits in its processor to process such a task. When the user terminal 130 issues an instruction to the server 110, the processor of the user terminal 130 may generate an electrical signal encoding the instruction. The processor of the user terminal 130 may then send the electrical signal to the output port. If the user terminal 130 communicates with the server 110 via a wired network, the output port may be physically connected to a cable, which further transmits electrical signals to the input port of the server 110 . If the user terminal 130 communicates with the server 110 via a wireless network, the output port of the user terminal 130 may be one or more antennas that convert electrical signals into electromagnetic signals. Similarly, storage module 140 may process tasks through the operation of logic circuits in its processor and receive instructions and/or information from server 110 via electrical or electromagnetic signals. Within an electronic device, such as user terminal 130, storage module 140 and/or server 110, when a processor processes an instruction, issues an instruction and/or performs an action, the instruction and/or action is performed through electrical signals. For example, when server 110 retrieves data from storage module 140, it can send electrical signals to a reading device of the storage medium, which can read structured or narrative information in the storage medium. The structured or descriptive information may be transmitted to the processor in the form of electrical signals via the bus of the electronic device. Here, an electrical signal may refer to one electrical signal, a series of electrical signals and/or at least two discrete electrical signals.

FIG. 2 is a schematic block diagram of an event attribute determination system according to some embodiments of the present application.

As shown in FIG. 2, in some embodiments, the event attribute determination system 200 may include: a work order acquiring module 210, configured to acquire work order information generated according to the event description of the target event, where the work order information at least includes Structural information and narrative information; the first feature extraction module 220 is used to obtain the first feature according to the narrative information; the second feature extraction module 230 is used to obtain the second feature according to the structured information; feature fusion The module 240 is configured to fuse the first feature and the second feature to obtain a fused feature; and a determination module 250 is configured to obtain the event attribute of the target event based on the fused feature.

In some embodiments, the work order acquisition module 210 may be used to perform step 301 . In some embodiments, the ticket information is generated by the customer service based on the event description and processing results of the event.

In some embodiments, the structured information in the work order information may include at least one of the following information: order number, license plate number, phone number related to the incident, whether to call the police, whether the police have filed a case, whether the user requests processing and the degree of urgency required for processing; the narrative information in the work order information includes at least one of the following information: a user's description of the event, a description of the police processing result, and a description of the customer service processing result. See step 301 for more related information.

In some embodiments, the first feature extraction module 220 may be used to perform step 303 . In some embodiments, the first feature extraction module may process the narrative information using a text transformation model to obtain the first feature. In some embodiments, the text transformation model may include at least one of the following deep learning models: Fast text, HAN, Text CNN, Transformer, LR, and XG Boost. In some embodiments, the method for training the text conversion model may include: acquiring a sample set, the sample set includes a plurality of pieces of work order information, taking narrative information in the work order information as input, and taking the information provided by customer service as input. The event attribute corresponding to the piece of work order information is used as an identifier to train the text conversion model. See step 303 for more related information.

In some embodiments, the second feature extraction module 230 may be used to perform step 305 . In some embodiments, the second feature extraction module may generate the second feature by processing structured information through an extraction model. In some embodiments, the extraction model may employ a rule-based and/or AC automaton-based approach to feature extraction. See step 305 for more related information.

In some embodiments, feature fusion module 240 may be used to perform step 307 . In some embodiments, the feature fusion module may directly concatenate the first feature and the second feature or process the first feature and the second feature through a set algorithm to generate a combined feature to obtain a fused feature. See step 307 for more related information.

In some embodiments, determination module 250 may be used to perform step 309 . In some embodiments, the determining module may process the fusion feature using a classification model to obtain a classification result of the event attribute. In some embodiments, the classification result of the event attribute obtained by the determining module may include at least one of the following: whether the event is safe, whether the event is accurate, whether the event is traceable, and whether the event is repeatable. In some embodiments, the classification model may include at least one of the following deep learning models: XG Boost, GBDT, Adaboost, Random Forest. In some embodiments, the method for training the classification model may include: acquiring a sample set, the sample set includes a plurality of pieces of work order information, using the fusion feature corresponding to the work order information as an input, and using the piece of information given by the customer service The event attribute corresponding to the work order information is used as an identifier to train the classification model. See step 309 for more related information.

It should be understood that the system and its modules shown in FIG. 2 may be implemented in various 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 part can be realized by using dedicated logic; the software part can be stored in a memory and executed 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, for example on a carrier medium such as a disk, CD or DVD-ROM, such as a read-only memory (firmware) ) or a data carrier such as an optical or electronic signal carrier. The system and its modules of the present application can not only be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. , can also be implemented by, for example, software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software (eg, firmware).

It should be noted that the above description of the system and its modules is only for the convenience of description, and does not limit the present application to the scope of the illustrated embodiments. It can be understood that for those skilled in the art, after understanding the principle of the system, various modules may be combined arbitrarily, or a subsystem may be formed to connect with other modules without departing from the principle. For example, in some embodiments, the work order acquisition module 210, the first feature extraction module 220, etc. disclosed in FIG. 2 may be different modules in a system, or may be a module that implements the functions of the above two or more modules . For another example, each module may share one storage module 140 , and each module may also have its own storage module 140 . Such deformations are all within the protection scope of the present application.

FIG. 3 is an exemplary flowchart according to some embodiments of the present application, which is used to illustrate the implementation steps of a method for determining an event attribute.

In other embodiments of the present application, a method for determining an event attribute is provided, and the method 300 may include the following steps:

Step 301: Obtain work order information generated according to the event description of the target event, where the work order information may at least include structured information and narrative information. In some embodiments, this step may be performed by the work order acquisition module 210 in the system 200 .

In some embodiments, the event description may include a user-provided description of the event. In some embodiments, the manner of event description provided by the user includes, but is not limited to, video, voice, text, and the like. The work order information can be obtained by processing the event description through the customer service. In some embodiments, the processing method described by the customer service processing event may be manual or semi-automatic, and semi-automation may refer to using various software and hardware to help the customer service process information.

In some embodiments, ticket information may include structured information and narrative information. Structured information can refer to information that can be decomposed into multiple interrelated components after analysis, and each component has a clear hierarchical structure. The use and maintenance of structured information can be managed through the database, and there are certain operating specifications. For example, structured information can be represented and stored using relational databases as data in two-dimensional form. The data is in row units, a row of data represents the information of an entity, and the attributes of each row of data are the same. The storage and arrangement of data is regular, which is convenient for subsequent query and modification operations. Structured information can be applied to, for example, enterprise ERP, financial systems, medical HIS databases, education cards, government administrative approvals, and other core databases. The storage scheme of structured information can include high-speed storage application requirements, data backup requirements, data sharing requirements, and data disaster recovery requirements.

In some embodiments, the structured information may include the order number, license plate number, phone number, whether to dial 110, whether the police have filed a case, etc. related to the incident, and some information that can be expressed as "yes or no", and can also refer to a hierarchical level information, such as whether the user requested processing and the urgency of the request. In some embodiments, the structured information may not conform to the form of a relational database or other data table, but contain relevant markup to separate semantic elements and to hierarchize records and fields. Therefore, structured information is also known as a self-describing structure. Structured information belongs to the same class of entities but can have different properties, and even if they are grouped together, the order of these properties does not matter. Common structured information includes XML and JSON, etc. Structured information can be applied to mail systems, WEB clusters, teaching resource libraries, data mining systems, file systems, and so on. The storage scheme of structured information can include basic storage requirements such as data storage, data backup, data sharing, and data archiving.

In some embodiments, descriptive information may refer to unstructured data whose data structure is irregular or incomplete, has no predefined data model, and is inconvenient to be represented by a two-dimensional logical table of a database. Narrative information can include office documents in all formats, text, pictures, various types of reports, images and audio/visual information, and so on. The formats of narrative information are very diverse, the standards are also diverse, and unstructured information is technically more difficult to standardize and understand than structured information. Therefore, storage, retrieval, publication and utilization require more intelligent information technology, such as mass storage, intelligent retrieval, knowledge mining, content protection, and value-added development and utilization of information. The storage scheme of narrative information can include data storage, data backup, and data sharing.

In some embodiments, the narrative information may refer to the user's description of the event recorded by the customer service, the description of the police handling result, the description of the handling result of the customer service, and the like. For example, the description of the user problem is "the car is scratched", the customer service process is "arrange a specialist and feedback to the traffic police", and the result is "user approval". Narrative information can be obtained by recording, saving videos, saving call records, etc.

In some embodiments, the ticket information may be generated by the customer service based on the event description and the processing result of the event. In some embodiments, the work order information may refer to information obtained after the user reports the incident to the customer service and the incident is preliminarily processed by the customer service. Ticket information is converted into structured information and/or narrative information in a manual or semi-automated manner by customer service agents. For example, the conversation between the user and the customer service is recorded and converted into structured information and/or narrative information according to the recording.

In some embodiments, the structured information in the work order information includes at least one of the following information: order number, license plate number, phone number, whether to call the police, whether the police file a case, whether the user requests processing, and The degree of urgency required for processing, etc.; the narrative information in the work order information includes at least one of the following information: the user's description of the event, the description of the police processing result, and the description of the customer service processing result.

As shown in Figure 4, the customer service agent presents the work order information as structured information and/or narrative information according to the event description. The narrative information obtained includes: 1. The description of the user's problem is: the user reported that he picked up the passenger by himself, and the passenger touched the door of the bicycle when he got into the car and drove the car, causing the door to be scratched. At present, he has asked the passenger for a fee of 100 yuan. Worrying about the passenger's complaint Own. 2. Customer service handling: I am sorry for the bad experience. We attach great importance to your feedback and will strengthen the control of passengers in the future. We will immediately report your questions to the relevant departments, and the relevant customer service specialists will be in 2 We will contact you every hour for processing, please keep the phone open. 3. Results: User approval. The obtained structured information may include information such as whether the user is advised to dial 110 as "no", the police processing result as "unknown", the user's complaint as "request to deal with", and whether the vehicle is maliciously damaged as "unknown".

Step 303: Obtain a first feature according to the descriptive information. In some embodiments, this step may be performed by the first feature extraction module 220 in the system 200 .

In some embodiments, obtaining the first feature according to the narrative information includes: using a text transformation model to process the narrative information to obtain the first feature. The text transformation model includes at least one of the following deep learning models: Fast text, HAN, Text CNN, Transformer, LR, and XG Boost, etc.

Fast text is a fast text classifier. Text classification refers to classifying documents into one or more categories, which can be rating scores, danger, urgency, harassment, etc. In order to build a classifier, label data needs to be obtained. Label data refers to data and the category (ie, identification or label) to which this data corresponds. For example, after being processed by the customer service, the customer service will manually give each work order a label of whether there is a danger. Among them, the data with category labels are historical data for a long time. Input a sequence of words (a piece of text or a sentence) into the Fast text model, and the model outputs the probability that the sequence of words belongs to different categories. The words and phrases in the sequence form a feature vector, and the feature vector is mapped to the hidden layer through a linear transformation, and the hidden layer is mapped to the label. Fast text categorizes narrative information in this way.

HAN (Hierarchy Attention Network) is a hierarchical model that uses the attention mechanism to model the importance of words and sentences in sentences and documents, respectively. The model corresponds to the hierarchical structure of documents: words form sentences, and sentences form documents, so the model is also divided into these two parts to construct a text vector representation. Secondly, different words and sentences have different amounts of information, and cannot be treated in a unified manner, so the Attention mechanism is introduced. In addition to improving the accuracy of the model, the introduction of the Attention mechanism can also analyze and visualize the importance of words and sentences, which enhances interpretability. HAN can include a word sequence encoder, a word-level attention layer, a sentence sequence encoder, a sentence-level attention layer, and so on.

Transformers can process all words or symbols in a sequence in parallel, while combining context with distant words using a self-attention mechanism. A self-attention mechanism refers to an attention mechanism that associates different positions of a single sequence. By processing all the words in parallel, and having each word notice the other words in the sentence over multiple processing steps.

LR (logistic regressive, logistic regression) or XG Boost (extreme gradient boosting, extreme gradient boosting) can assign a specific weight value to each word or word of the text of the training sample according to the corresponding classification. For example, in the process of using training samples to train XG Boost, XG Boost can give weights of words or words in the samples to indicate how important each word or word is to model training.

In some embodiments, the method for training the text conversion model includes: acquiring a sample set, the sample set includes multiple pieces of work order information, taking narrative information in the work order information as input, and using the The event attribute corresponding to the piece of work order information is used as an identifier to train the text conversion model. The event attributes given by the customer service may include at least one of the following: whether the event is safe, whether the event is accurate, whether the event is traceable, whether the event is repeatable, whether the event is urgent, and whether the event can be warned. In Fast text, sample sets are labeled narrative information. The sample set can be descriptive information with very large classification categories and enough data set to avoid overfitting. Input narrative information into the model, the model can output the probability that the narrative information belongs to different categories. The model can also output the most probable label category corresponding to the descriptive information. In HAN, sample sets are labeled narrative information. The processor can convert the labeled narrative information into a sequence of word vectors by preprocessing it, and input the sequence of word vectors into the model. Preprocessing refers to the initial processing of labeled narrative information. In some embodiments, preprocessing may include unifying lower case, removing garbled characters, removing abbreviations and numbers, removing stop words, and the like. Stop words refer to some frequently occurring words that have little effect on classification. Stop words that appear in the text can be removed according to the stop word list. During the training process, the model trains the weight of the word in the sentence, the weight of the training sentence in the document, and the vector representation of the output text. In some embodiments, the entire text can also be classified using a softmax classifier. The softmax classifier can output the probability to which the corresponding information belongs respectively. The model can output the label category with the highest probability corresponding to the descriptive information. In Transformer, sample sets are labeled narrative information. Enter narrative information into the model. The output narrative information corresponds to the probabilities of different classes. Output the label category with the highest probability corresponding to the descriptive information. In logistic regression (LR) or XG Boost, the sample set is labeled narrative information. Input the narrative information into the model, assign a specific weight value to each character or word of the training samples in the sample set, and output the label category with the largest weight corresponding to the narrative information.

Step 305, obtaining a second feature according to the structured information. In some embodiments, this step may be performed by the second feature extraction module 230 in the system 200 .

In some embodiments, obtaining the second feature according to the structured information includes: processing the structured information through an extraction model to generate the second feature. For example, "Whether users are advised to dial 110". If it is "No", the extraction method is: "None"→feature_110_0→The corresponding dimension in feature_110_0 is 1; if it is "Yes", the extraction method is: "Yes"→feature_110_1→The corresponding dimension in feature_110_1 is 1. In some embodiments, the extraction model may employ a rule-based and/or AC automaton-based approach to feature extraction. An AC automaton is a string search algorithm for matching substrings in a finite set of "dictionaries" within an input string of strings. Dictionary refers to a collection of some elements, some elements can include words and sentences in the work order information, for example, "user processing", user processing corresponds to feature_120, if it is "request processing", the extraction method is: "yes" → feature_120_1 →The corresponding dimension in feature_120_1 is 1; if it is "no processing required", the extraction method is "None" →feature_120_0 → The corresponding dimension in feature_120_0 is 1. The AC automaton can match all strings at the same time.

Step 307, fuse the first feature and the second feature to obtain a fused feature. In some embodiments, this step may be performed by feature fusion module 240 in system 200 .

In some embodiments, the method for fusing the first feature and the second feature to obtain a fused feature includes: directly splicing or processing the first feature and the second feature through a set algorithm to generate a combined feature. In some embodiments, the method for generating combined features includes, but is not limited to, feature combination, feature dimensionality reduction, feature intersection, and the like. Feature combination refers to obtaining a new feature through some linear or nonlinear superposition of features. Common feature combination methods can be Cartesian product methods. For example, A={0,1}, B={0,1}, then A×B={(0,0),(0,1),(1,0),(1,1)}. The feature combination method can also adopt the decision tree + LR method. In some embodiments, a gradient boosted tree+logistic regression (GBDT+LR) approach can be used. Gradient boosting tree + logistic regression (GBDT + LR) is an automatic feature extraction method. GBDT is a gradient boosting decision tree. First, a decision tree will be constructed. First, a decision tree will be constructed on the residuals of the existing model and the actual sample output, and it will be iterated continuously. Each iteration will generate a classification with a larger gain. Therefore, the resulting feature space is as large as the number of leaf nodes in the decision tree constructed by GBDT, and this feature is used as the input of the LR model. A decision tree is represented as a tree of nodes, where each node makes a binary decision based on the characteristics of the data, and each leaf node of the tree contains a prediction.

The method of feature combination can also be implemented based on the principle of principal component analysis (PCA), feature dimension reduction of singular value decomposition (SVD), and feature intersection based on factorization machine (FM, Factorization Machine). PCA refers to the synthesis of multiple indicators into a few independent comprehensive indicators (ie principal components), each of which can reflect most of the information of the original variables, and the information contained does not repeat each other. The basic principle of Principal Component Analysis (PCA) is to project the sample data in a matrix into a new space. For a matrix, diagonalizing it, that is, the process of generating eigenroots and eigenvectors, is also the process of projecting it on a standard orthonormal basis. The eigendirection reflects the distribution state of the information, and the eigenvalue defines the length of the direction. For example, the larger the eigenvalue and the longer the length of the feature direction, the more information of the original data in this direction. Singular value decomposition (SVD) is an important matrix decomposition method. SVD can be applied to numerical data types, and can use smaller data sets to represent original data sets. The factorization machine (FM) can solve the problem of feature combination in the case of sparse data.

Step 309, based on the fusion feature, obtain the event attribute of the target event. In some embodiments, this step may be performed by the determination module 250 in the system 200 .

In some embodiments, the obtaining the classification result of the event attribute based on the fusion feature includes: using a classification model to process the fusion feature to obtain the classification result of the event attribute. In some embodiments, the classification result of the event attribute includes at least one of the following: whether the event is safe, whether the event is accurate, whether the event is traceable, and whether the event is repeatable. In some embodiments, the event attribute classification may not be limited to the above four, and may also include whether the event is urgent, whether the event can be forewarned, and the like. In some embodiments, the output result of the classification result may be a binary classification result, that is, only two types of results, safe or unsafe, may be output, or it may be a multi-classification result, that is, a probability that the output event is a safety event. For example, the binary classification result is specifically expressed as safe output 1 and unsafe output 0. For another example, the multi-classification result is specifically expressed as the probability output of safety events is 0.8, the event probability of unsafe events is 0.1, and the event probability of disturbance events is 0.1. The classification model includes at least one of the following deep learning models: XG Boost, GBDT, Adaboost, and random forest.

Adaboost is a boosting tree, which trains different weak classifiers for the same training set, and builds a strong classifier through Adaboost. Boosting tree refers to the promotion of a weak learning algorithm to a strong learning algorithm. A classifier is a function or model that can map data records in the database to one of a given category, which can be applied to data prediction. Classifier is a general term for the methods of classifying samples in data mining, at least including decision trees, logistic regression, naive Bayes, neural networks and other algorithms. There are two weights in the Adaboost algorithm, one is the weight of the data, and the other is the weight of the weak classifier. Among them, the weight of the data is mainly used for the weak classifier to find the decision point with the smallest classification error, and after finding the minimum classification error, the weight of the weak classifier is calculated. have greater weight.

Random forest (RF, Random forests) refers to a classifier that uses multiple trees to train and predict samples. The categories it outputs are determined by the mode of the categories output by the individual trees.

In some embodiments, the method for training the classification model includes: acquiring a sample set, the sample set includes a plurality of pieces of work order information, using the fusion feature corresponding to the work order information as an input, and using the piece of work given by the customer service The event attribute corresponding to the single message is used as an identifier to train the classification model.

In the Adaboost model training process, the weak classifier is trained first, and the trained weak classifier will participate in the next iteration. In the Nth iteration, there are a total of N weak classifiers, of which N-1 weak classifiers are previously trained, and their various parameters are not changed. This time, the Nth classifier is trained. The relationship between the weak classifiers is that the Nth weak classifier is more likely to classify the data that the first N-1 weak classifiers are not paired with, and the final classification output depends on the combined effect of the N classifiers. Adaboost generally uses a single-layer decision tree as its weak classifier. It should be noted that even if the sample set has multi-dimensional features, a single-level decision tree can only choose one of the dimensions to make decisions. In a neural network, the classified data is converted into a vector form and input to the neural network for classification.

It should be noted that the above description of the related processes is only for example and description, and does not limit the scope of application of the present application. For those skilled in the art, various modifications and changes can be made to the procedures under the guidance of the present application. However, such corrections and changes are still within the scope of this application. For example, Decision Tree+Logistic Regression can be implemented in ways other than LBDT+LR (eg, RF+LR, XG Boost+LR, etc.).

In other embodiments of the present application, an event attribute determination apparatus is provided, comprising at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute the At least some of the computer instructions to implement the operations as described above.

In still other embodiments of the present application, a computer-readable storage medium for event attribute determination is provided, the storage medium stores computer instructions that, when executed by a processor, implement the operations described above .

It should be noted that the above description is only for the convenience of description, and does not limit the present application to the scope of the illustrated embodiments. It can be understood that, for those skilled in the art, after understanding the principle of the present application, various modifications and changes in form and details can be made to the above process without departing from the principle. However, these changes and modifications do not depart from the scope of this application.

The possible beneficial effects of the embodiments of the present application include, but are not limited to: (1) the work order information contains both narrative information and structured information, and to a certain extent, the narrative information can also reflect the processing results of the customer service on the event, The event attributes are obtained by combining the processing results of the two types of information, thereby improving the accuracy of event attribute determination; (2) The narrative information and structured information are processed separately and input into the model, and there is no need to perform correlation conversion on the two types of information, which improves the processing efficiency. efficiency.

It should be noted that different embodiments may have different beneficial effects, and in different embodiments, the possible beneficial effects may be any one or a combination of the above, or any other possible beneficial effects.

The foregoing describes the present application and/or some other examples. According to the above content, the present application can also make different modifications. The subject matter disclosed in this application can be implemented in different forms and examples, and the application can be used in a wide variety of applications. All applications, modifications and changes claimed in the following claims are within the scope of this application.

Meanwhile, the present application uses specific words to describe the embodiments of the present application. Such as "one embodiment," "an embodiment," and/or "some embodiments" means a certain feature, structure, or characteristic associated with at least one embodiment of the present application. Accordingly, it should be emphasized and noted that two or more references to "an embodiment," or "one embodiment," or "an alternative embodiment," or "another implementation," in various places throughout this specification are example" or "another embodiment" are not necessarily referring to the same embodiment. Furthermore, certain features, structures or characteristics of the one or more embodiments of the present application may be combined as appropriate.

It will be understood by those skilled in the art that various modifications and improvements may occur to the content disclosed in this application. For example, the various system components described above are implemented by hardware devices, but may also be implemented by software-only solutions. For example: installing the system on an existing server. Additionally, the provision of the location information disclosed herein may be implemented by a firmware, a firmware/software combination, a firmware/hardware combination, or a hardware/firmware/software combination.

All or part of the software may sometimes communicate over a network, such as the Internet or other communication network. Such communications enable the loading of software from one computer device or processor to another. For example: a hardware platform loaded from a management server or host computer of a radiation therapy system to a computer environment, or other computer environment implementing the system, or a system of similar functionality related to providing the information needed to determine the target structure parameters of the wheelchair. Therefore, another medium capable of transmitting software elements can also be used as a physical connection between local devices, such as light waves, radio waves, electromagnetic waves, etc., through cables, optical cables or air. The physical medium used for the carrier wave, such as a cable, wireless connection, or fiber optic cable, etc., can also be considered to be the medium that carries the software. Unless the usage herein is limited to tangible "storage" media, other terms referring to computer or machine "readable media" refer to media that participate in the execution of any instructions by a processor.

The computer program coding required for the operation of the various parts of this application may be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python etc., conventional procedural programming languages such as C language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages, etc. The program code may run entirely on the user's computer, or as a stand-alone software package on the user's computer, or partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the latter case, the remote computer may be connected to the user's computer through any network, such as a local area network (LAN) or wide area network (WAN), or to an external computer (eg, through the Internet), or in a cloud computing environment, or as a Service usage such as Software as a Service (SaaS).

Furthermore, unless explicitly stated in the claims, the order of processing elements and sequences described in the present application, the use of numbers and letters, or the use of other names are not intended to limit the order of the procedures and methods of the present application. While the foregoing disclosure discusses by way of various examples some embodiments of the invention that are presently believed to be useful, it is to be understood that such details are for purposes of illustration only and that the appended claims are not limited to the disclosed embodiments, but rather The requirements are intended to cover all modifications and equivalent combinations falling within the spirit and scope of the embodiments of the present application. 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 systems on existing servers or mobile devices.

Similarly, it should be noted that, in order to simplify the expressions disclosed in the present application and thus help the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of the present application, various features are sometimes combined into one embodiment, in the drawings or descriptions thereof. However, this method of disclosure does not imply that the subject matter of the application requires more features than those mentioned in the claims. Indeed, there are fewer features of an embodiment than all of the features of a single embodiment disclosed above.

In some embodiments, numbers describing properties and quantities are used, and it should be understood that such numbers used to describe the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples . Unless stated otherwise, "about", "approximately" or "substantially" means that a variation of ±20% is allowed for the stated number. Accordingly, in some embodiments, the numerical parameters set forth in the specification and claims are approximations that can vary depending upon the desired characteristics of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and use a general digit reservation method. Notwithstanding that the numerical fields and parameters used in some embodiments of the present application to confirm the breadth of their ranges are approximations, in particular embodiments such numerical values are set as precisely as practicable.

Each patent, patent application, patent application publication, and other material, such as article, book, specification, publication, document, article, etc., cited in this application is hereby incorporated by reference in its entirety. Application history documents that are inconsistent with or conflict with the content of this application are excluded, as are documents (currently or hereafter appended to this application) that limit the broadest scope of the claims of this application. It should be noted that, if there is any inconsistency or conflict between the descriptions, definitions and/or terms used in the attached materials of this application and the content of this application, the descriptions, definitions and/or terms used in this application shall prevail .

Finally, it should be understood that the embodiments described in the present application are only used to illustrate the principles of the embodiments of the present application. Other variations are also possible within the scope of this application. Accordingly, by way of example and not limitation, alternative configurations of embodiments of the present application may be considered consistent with the teachings of the present application. Accordingly, the embodiments of the present application are not limited to the embodiments explicitly introduced and described in the present application.

Claims (15)

1. A method for event attribute determination, the method being implemented by at least one processor, the method comprising:

Acquiring work order information generated according to event description of a target event, wherein the work order information at least comprises structural information and narrative information;

obtaining a first characteristic according to the narrative information;

obtaining a second characteristic according to the structural information;

fusing the first feature and the second feature to obtain a fused feature;

and obtaining the event attribute of the target event based on the fusion characteristic.

2. The method of claim 1, wherein the work order information is generated by a customer service based on the event description and/or results of processing an event.

3. The method of claim 2, wherein the structured information in the work order information comprises at least one of: an order number, license plate number, phone number, whether to alarm, whether to set up a case for the police, whether to request processing by the user, and the urgency of the processing required; the narrative information in the work order information at least comprises at least one of the following information: the description of the user to the event, the description of the police processing result and the description of the customer service processing result.

4. The method of claim 2, wherein said deriving a first feature from said narrative information comprises: processing the narrative information by using a text conversion model to obtain the first characteristic; the first feature reflects a first determination result of an event attribute of the target event.

5. The method of claim 4, wherein the text conversion model comprises at least one of the following deep learning models: fast Text, HAN, Text CNN, Transformer, LR, and XG Boost.

6. The method of claim 4 or 5, wherein the method of training the text conversion model comprises:

acquiring a sample set, wherein the sample set comprises a plurality of pieces of work order information;

and training the text conversion model by taking the narrative information in the work order information as input and taking the event attribute corresponding to the work order information given by the customer service as an identifier.

7. The method of claim 1, wherein the deriving the second feature from the structured information comprises: the second feature is generated by extracting model processing structured information.

8. The method of claim 7, wherein the extraction model performs feature extraction using a rule and/or AC automata based approach.

9. The method of claim 1, wherein fusing the first and second features to obtain a fused feature comprises: and processing the first feature and the second feature directly or through a set algorithm to generate a combined feature.

10. The method of claim 1, wherein the deriving the event attribute of the target event based on the fused feature comprises: processing the fusion characteristics by using a classification model to obtain a classification result of the event attributes; the classification result of the event attribute comprises at least one of the following: whether the event is safe, whether the event is accurate, whether the event can be traced, and whether the event can be repeated.

11. The method of claim 10, wherein the classification model comprises at least one of the following deep learning models: XG Boost, GBDT, Adaboost, random forest.

12. The method of claim 10 or 11, wherein the method of training the classification model comprises:

acquiring a sample set, wherein the sample set comprises a plurality of pieces of work order information;

and taking the fusion characteristics corresponding to the work order information as input, taking the event attribute corresponding to the work order information given by the customer service as an identifier, and training the classification model.

13. An event attribute determination system, comprising:

the work order obtaining module is used for obtaining work order information generated according to the event description of the target event, and the work order information at least comprises structural information and narrative information;

The first feature extraction module is used for obtaining a first feature according to the narrative information;

the second feature extraction module is used for obtaining a second feature according to the structural information;

the feature fusion module is used for fusing the first feature and the second feature to obtain a fusion feature; and

and the determining module is used for obtaining the event attribute of the target event based on the fusion characteristic.

14. An event attribute determination 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 implement the operations of any of claims 1 to 12.

15. A computer-readable storage medium for event attribute determination, the storage medium storing computer instructions which, when executed by a processor, perform operations according to any one of claims 1 to 12.

Images