Order completion method, system and device and medium

Abstract

The invention relates to an order completion method, system and device and a medium. The method comprises the steps of obtaining an order of a user; on the basis of a keyword matching method, according to the content of the order, identifying an order type; according to the order type, detecting a key item of the order, and when it is detected that the key item is incomplete, triggering an outbound system; wherein the outbound system is used for establishing dialogue contact with a user; receiving a user reply returned by the outbound system, and inputting the received user reply into a large language model for intention recognition to recognize necessary information; and the necessary information is complemented to the corresponding incomplete key items, and complete order information is generated. According to the method, necessary information provided by the user can be efficiently identified and extracted, so that missing key items are complemented, and the integrity of order information is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of order management, and in particular to an order completion method, system, device and medium. Background Art

[0002] In modern freight logistics, transport ordering systems allow users to post their cargo transportation needs online, and drivers use this information to decide whether to accept the order. However, the information provided by users often lacks key details such as cargo size, weight, or special transportation requirements, which limits the efficiency of order matching.

[0003] Existing technologies have attempted to address information asymmetry, such as manual information entry via phone calls, text messages, or mobile applications. However, these methods are often time-consuming and rely heavily on manual intervention, failing to guarantee the timeliness and accuracy of information collection. While some systems use technologies based on traditional natural language processing (NLP) algorithms (such as BERT) to automate this process, they still have limitations in understanding context and tracking long conversations. Therefore, there is a need to provide an order completion method, system, device, and medium. Summary of the Invention

[0004] In view of the shortcomings of the prior art described above, the purpose of the present invention is to provide an order completion method, system, device and medium, which improves the problem that the prior art has difficulty in dynamically understanding the user's answer content, resulting in low accuracy in order completion.

[0005] To achieve the above-mentioned purpose and other related purposes, the present invention provides an order completion method, comprising: obtaining a user's order; identifying the order type according to the content of the order based on a keyword matching method; detecting key items of the order according to the order type, and triggering an outbound call system when it is detected that the key items are incomplete; wherein the outbound call system is used to establish a dialogue connection with the user; receiving a user reply returned by the outbound call system, and inputting the received user reply into a large language model for intent recognition to identify necessary information; and completing the necessary information to the corresponding incomplete key items to generate complete order information.

[0006] In one embodiment of the present invention, the keyword matching method is based on the content of the order to identify the order type, including: extracting text information from the order; performing text cleaning and standardization processing on the text information to obtain preprocessed text information; based on a preset word segmentation algorithm, performing word segmentation processing on the preprocessed text information to obtain a text word set; based on a keyword library, extracting keywords from the text word set; according to the type of the keyword, determining the corresponding order type; wherein each type of keyword corresponds to an order type.

[0007] In one embodiment of the present invention, the method of receiving the user reply returned by the outbound call system and inputting the received user reply into the large language model for intent recognition to identify the necessary information therein includes: receiving the user reply, calling the interface of the large language model, inputting the user reply into the large language model, and judging whether the user reply contains the necessary information: if it does, extracting the necessary information; if not, generating a new guiding question based on the user reply, feeding back the new guiding question to the outbound call system, and continuously monitoring the outbound call system within a preset time period.

[0008] In one embodiment of the present invention, the process of generating the new guiding question includes: extracting the reply keywords of the user's reply; determining whether the extracted reply keywords match the preset question template; wherein each question in the question template corresponds to at least one reply keyword; if matched, calling the corresponding question as the new guiding question; if not matched, inputting the user's reply into the pre-trained dialogue generation language model to generate a new guiding question.

[0009] In one embodiment of the present invention, the completion of necessary information to the corresponding incomplete key items to generate complete order information includes: inputting the necessary information into a pre-trained information type recognition model to identify the information type of the necessary information; wherein, the information type recognition model is a natural language processing model; based on the preset mapping relationship between information type and key items, filling in the corresponding necessary information in the preset position of each incomplete key item to generate complete order information.

[0010] In one embodiment of the present invention, after supplementing the key items according to the necessary information and generating complete order information, it also includes: matching drivers according to the order information based on a preset driver matching algorithm, determining the drivers that match the order and recommending them.

[0011] In one embodiment of the present invention, the driver matching algorithm based on a preset method matches drivers according to the order information, determines drivers that match the order, and recommends them, including: extracting key items from the order information; screening an initial set of drivers from a driver database based on the key items; calculating the comprehensive matching degree of each initial driver in the initial set of drivers based on multiple preset matching factors, and screening out the driver with the highest comprehensive matching degree as the driver that matches the order and recommending him.

[0012] In one embodiment of the present invention, an order completion system is also provided, which includes: an order acquisition module for acquiring a user's order; a type identification module for identifying the order type according to the content of the order based on a keyword matching method; a key item detection module for detecting the key items of the order according to the order type, and triggering an outbound call system when it is detected that the key items are incomplete; wherein the outbound call system is used to contact the user; an intent recognition module for receiving the user's reply returned by the outbound call system, and inputting the received user reply into a large language model for intent recognition to identify the necessary information therein; an information supplement module for supplementing the key items according to the necessary information to generate complete order information.

[0013] In one embodiment of the present invention, an electronic device is also provided, including: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the electronic device implements any of the order completion methods described above.

[0014] In one embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor of a computer, the computer executes any of the above-mentioned order completion methods.

[0015] As described above, the order completion method, system, device and medium of the present invention have the following beneficial effects: the order type can be accurately identified according to the order content through the keyword matching method, and the type can be used to automatically detect whether the key items required to be filled in the order are complete. When the filling is incomplete, the outbound call system is automatically triggered to communicate with the user, so that the reply information provided by the user can be obtained more quickly, improving the information delay problem caused by manual return visits. By inputting the user's reply into the large language model for intent recognition, the necessary information provided by the user can be efficiently identified and extracted, thereby completing the missing key items and ensuring the integrity of the order information. Through this automated method, with the help of the excellent context understanding ability of the large language model, the user's reply information can be understood more accurately, and the necessary information provided by the user can be identified and extracted, which greatly improves the automation level and accuracy of information collection, reduces manual intervention, reduces operating costs and shortens processing time. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a method for completing an order provided in an embodiment of the present invention;

[0017] Figure 2 Shown is a structural block diagram of an order completion system provided by an embodiment of the present invention;

[0018] Figure 3 Shown is a structural schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following describes the embodiments of the present invention through specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0020] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0021] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the embodiments of the present invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring the embodiments of the present invention.

[0022] The inventors discovered that on existing freight service platforms, users often fail to provide comprehensive and detailed cargo information, leading to uncertainty among drivers regarding order details and requirements before accepting them. Furthermore, traditional rule-based NLP systems often suffer from low outbound call efficiency and inaccurate information due to limitations in contextual understanding.

[0023] The present invention provides an order completion method, which can accurately identify the order type according to the order content through a keyword matching method, and automatically detect whether the key items required to be filled in the order are complete through the type. When the filling is incomplete, the outbound call system is automatically triggered to communicate with the user, so that the reply information provided by the user can be obtained more quickly, improving the information delay problem caused by manual return visits. By inputting the user's reply into a large language model for intent recognition, the necessary information provided by the user can be efficiently identified and extracted, thereby completing the missing key items and ensuring the integrity of the order information. Through this automated method, the automation level and accuracy of information collection are greatly improved, manual intervention is reduced, operating costs are reduced, and processing time is shortened.

[0024] See Figure 1 , the order completion method includes the following steps:

[0025] S1. Get the user's order.

[0026] Orders can be ride orders, freight orders, etc. For ease of description, the following takes freight orders as an example. Users submit freight demands through online platforms such as web pages or mobile applications, and fill in the corresponding order information. Among them, order information includes but is not limited to information such as cargo type, origin and destination, estimated transportation time, cargo size, cargo weight, and transportation conditions. However, when filling out an order, the user may not provide all the necessary information required, such as cargo weight, cargo transportation requirements, etc., resulting in missing relevant information and inability to reasonably dispatch orders. Therefore, the freight platform needs to identify the missing information and fill it in to form a complete order.

[0027] S2. Based on a keyword matching method, the order type is identified according to the content of the order.

[0028] After the freight platform receives the order submitted by the user, it extracts a series of text words from the order and identifies the keywords in the text words by matching them with the preset keyword library. Since each order type is pre-set with a specific set of keywords, when certain keywords exist in the order, the specific type of order can be identified based on the keywords. For example, if the keyword of the order is "low temperature transportation", the corresponding order type is "cold chain transportation". It should be noted that when the order uploaded by the user is in the form of an image, the text in the image can be extracted through the OCR method. When the user submits the order through a web page or mobile application, since the order is submitted in the form of structured form data, the corresponding text information can be obtained by extracting the form field.

[0029] In one embodiment of the present invention, the keyword matching method is used to identify the order type according to the content of the order, including:

[0030] extracting text information from the order;

[0031] Performing text cleaning and standardization processing on the text information to obtain preprocessed text information;

[0032] Based on a preset word segmentation algorithm, the preprocessed text information is segmented to obtain a text word set;

[0033] Extracting keywords from the text word set based on a keyword library;

[0034] According to the type of the keyword, the corresponding order type is determined; wherein each type of keyword corresponds to one order type.

[0035] Since orders include not only text but also images and other information, it is necessary to extract text information from user-submitted orders for subsequent analysis. Considering that original text often contains special symbols and spelling errors, the extracted text information needs to be cleaned and standardized to improve subsequent recognition accuracy. Text cleaning includes, but is not limited to, removing irrelevant characters, deleting stop words, and correcting spelling errors. Standardization includes, but is not limited to, standardizing capitalization and time formats. After these text cleaning and standardization processes, preprocessed text information is obtained. The preprocessed text information is segmented using a word segmentation algorithm, thereby dividing a continuous text segment into several words to form a text word set. It is understood that the word segmentation algorithm can be any method, such as rule-based, statistical, or deep learning-based, and is not limited here. Each text word in the text word set is compared one by one with a word in the keyword library to determine whether the two match. The matching text words are then used as keywords for the order. Since each order type corresponds to a specific keyword type, and there can be multiple keywords of the same type, there are multiple keywords. For example, keywords for "cold chain transportation" may include: refrigeration, low temperature, refrigerator, temperature control, etc. By identifying the type of keywords, the corresponding order type can be determined.

[0036] S3. Detect key items of the order according to the order type, and trigger an outbound call system when it is detected that the key items are incomplete; wherein the outbound call system is used to establish a dialogue connection with the user.

[0037] Based on the order information submitted by the user, the freight platform determines the order type through keyword matching. It then creates a list of key items corresponding to the order type and checks each key item in the user's order to ensure that all relevant key items are complete. If any key items are missing or incomplete, the freight platform automatically triggers an outbound call system. The outbound call system connects with the user via phone, text message, or other communication methods and intelligently generates guiding questions based on the missing key items. Preferably, to improve efficiency, the outbound call system conducts an automated conversation with the user over the phone. Key items refer to the fields required to complete the order, and required information refers to the content of the fields corresponding to the key items, i.e., the specific content of the key items. For example, the key item is "cargo weight" and the required information is "30 kilograms." During the phone call, the outbound call system intelligently generates questions based on the missing key items. For example, if the key item "cargo weight" is missing in the order, the outbound call system automatically asks the user, "What is the cargo weight?" to guide the user to provide the missing information and complete the key items in the order.

[0038] S4. Receive the user response returned by the outbound call system, and input the received user response into the large language model for intent recognition to identify necessary information.

[0039] After the outbound call system asks the user for the missing necessary information, it will receive the user's reply information and determine the type of the user's reply information: if the reply information is a text message replied by the user through SMS or other means, the reply information will remain unchanged and no further processing is required; if the reply information is a voice message replied by the user through a phone call or voice message, the voice message will be converted into a text message based on the voice recognition algorithm. The textual reply information and the prompt are input into the large language model for intent recognition to identify the necessary information. Among them, the prompt is a preset short description used to help the large language model better understand the contextual information and improve the accuracy of the final reply. For example, the prompt is "The following is the user's reply, please identify the necessary information." Among them, the large language model includes but is not limited to the GPT series model, T5 model or BLOOM, etc.

[0040] In one embodiment of the present invention, the receiving of the user response returned by the outbound call system and inputting the received user response into the large language model for intent recognition to identify necessary information therein include:

[0041] Receive the user's reply, call the interface of the large language model, input the user's reply into the large language model, and determine whether the user's reply contains necessary information:

[0042] If included, the necessary information is extracted;

[0043] If not, a new guiding question is generated based on the user's answer, the new guiding question is fed back to the outbound calling system, and the outbound calling system is continuously monitored within a preset time period.

[0044] After receiving the user's response, the freight platform invokes the large language model's interface and inputs the user's response into the large language model for processing. By identifying the user's response's intent, the large language model can determine whether it contains the necessary information. If the user's response contains the necessary information, the large language model extracts it and marks it as the corresponding field, such as cargo weight, dimensions, and delivery address, to facilitate subsequent order data replenishment and processing. Conversely, if the user's response does not include the necessary information, the large language model uses the outbound call system to send new guiding questions to the user based on the current conversation content and the missing necessary information. These questions are more targeted and aim to guide the user to provide more precise information. For example, if the user failed to provide the specific cargo weight in the previous conversation, the new guiding question might be "Please provide the specific weight of the cargo." After asking the user a question, the outbound call system waits for a preset response time. If no response is received within the preset time, it indicates that the user has not responded in a timely manner. In this case, the freight platform will take further action based on pre-set strategies, such as re-calling or sending a text message reminder. Conversely, if a user response is received within the preset timeframe, the new response is fed into the large language model for further analysis to determine whether it contains the missing necessary information. If so, the order is updated and the data is supplemented. Otherwise, a new guiding question is generated. The collaborative work of the large language model and the outbound call system effectively improves the efficiency of supplementing missing information, reduces manual intervention, and enhances the user experience.

[0045] In one embodiment of the present invention, the process of generating the new guiding question includes:

[0046] Extracting reply keywords of the user's reply;

[0047] Determining whether the extracted answer keywords match a preset question template; wherein each question in the question template corresponds to at least one answer keyword;

[0048] If there is a match, the corresponding question is called as the new guiding question;

[0049] If there is no match, the user's response is input into a pre-trained dialogue generation language model to generate a new guiding question.

[0050] When the freight platform receives a user's response message through the outbound call system, it can extract relevant response keywords using keyword extraction. Response keywords include, but are not limited to, cargo dimensions, cargo weight, and transportation conditions. It should be noted that keyword extraction methods include, but are not limited to, inputting the response message into a natural language processing model to obtain keywords, or extracting keywords based on word frequency. The response keywords are then matched to a pre-set template, where each question has at least one response keyword. For example, for a question about cargo dimensions, the response keywords might be "size," "size," "length, width, and height," etc. If the extracted response keywords match one of the questions, that question is sent to the user via the outbound call system as a guiding question. Conversely, if the extracted response keywords do not match any of the questions, the user's response is input into a large language model, which, by understanding the user's current response, generates a more accurate guiding question. This combination of a large language model and response templates significantly reduces computing resource consumption and quickly identifies necessary information for standardized tasks using pre-set templates. For more complex and changeable scenarios, the large language model can understand more complex user intentions based on the context and generate personalized guidance questions, thereby improving the user experience and avoiding situations where templates are not applicable.

[0051] S5. Complete the necessary information to the corresponding incomplete key items to generate complete order information.

[0052] After receiving the user's response from the outbound call system, the necessary information is extracted from the user's response through a large language model, and the key items corresponding to each necessary information are confirmed. Since the key item represents a specific data field in the order, by matching these necessary information with the corresponding fields in the order, if the matched fields are missing or incomplete, the corresponding necessary information will be supplemented to the corresponding fields. For example, if the order lacks the weight information of the goods, the outbound call system obtains the relevant necessary information, such as "50 kilograms", by asking and answering the user. At this time, the necessary information will be filled in the "Goods Weight" field of the order. After completing the supplement of all necessary information, the freight platform will generate a complete order information.

[0053] In one embodiment of the present invention, the step of completing the necessary information to the corresponding incomplete key items to generate complete order information includes:

[0054] Inputting the necessary information into a pre-trained information type recognition model to identify the information type of the necessary information; wherein the information type recognition model is a natural language processing model;

[0055] Based on the mapping relationship between the preset information type and key items, the corresponding necessary information is filled in the preset position of each incomplete key item to generate complete order information.

[0056] The freight platform extracts necessary information from the user's response by calling a large language model, and inputs the extracted necessary information into a pre-trained information type recognition model, and determines its information type by analyzing the characteristics of the necessary information. It can be understood that the necessary information types include but are not limited to cargo weight, cargo size, transportation conditions, etc. Among them, the information type recognition model includes but is not limited to time series natural language processing models such as BERT and Transformer series. According to the mapping relationship between the preset information type and key items, the identified necessary information is filled in the corresponding order key items to form a complete order. For example, if the information type "500kg" is recognized as "cargo weight", the freight platform will fill this information into the "cargo weight" field of the order.

[0057] In one embodiment of the present invention, after completing the order information by supplementing the key items with the necessary information, the process further includes: matching drivers based on the order information using a preset driver matching algorithm, identifying and recommending drivers who match the order. After completing the order information, the freight platform intelligently identifies and recommends the best drivers to the user based on the cargo information and the driver's historical performance, preferences, geographic location, and other parameters.

[0058] In one embodiment of the present invention, the driver matching based on the order information based on a preset driver matching algorithm to determine a driver matching the order includes: extracting key items from the order information;

[0059] According to the key items, an initial set of drivers is selected from the driver database;

[0060] Based on a plurality of preset matching factors, the comprehensive matching degree of each initial driver in the initial driver set is calculated, and the driver with the highest comprehensive matching degree is screened out and recommended as the driver matching the order.

[0061] After the freight platform completes the necessary information in the user's order, it will screen drivers from its driver database based on key items in the order (such as cargo type, origin and destination, transportation conditions, etc.) to initially meet the requirements. It is understood that if a driver meets only one of the key items, they will be screened out. For example, drivers within N kilometers of the destination or drivers capable of cold chain transportation are screened out. Based on multiple preset matching factors, these matching factors are weighted and summed according to their corresponding preset weights. An overall matching score is calculated for each selected driver. The driver with the highest overall matching score is selected as the driver matching the order and recommended to the user. For example, geographic location matching and vehicle matching are given higher weights, while the driver's historical performance is given lower weights. By setting different weights, an overall matching score for each driver is ultimately determined. Furthermore, once the order is accepted by the driver, the freight platform will record the corresponding driver and order information for subsequent optimization. After the order is completed, feedback from the driver and user is collected to further improve the matching algorithm and the outbound call interaction experience.

[0062] See Figure 2 , the order completion system 100 includes: an order acquisition module 110, a type identification module 120, a key item detection module 130, an intention identification module 140 and an information supplementation module 150. The above-mentioned order acquisition module 110 is used to obtain the user's order. The type identification module 120 is used to identify the order type according to the content of the order based on the keyword matching method. The key item detection module 130 is used to detect the key items of the order according to the order type, and trigger the outbound call system when it is detected that the key items are incomplete; wherein, the outbound call system is used to contact the user. The intention identification module 140 is used to receive the user's reply returned by the outbound call system, and input the received user reply into the large language model for intention identification to identify the necessary information therein. The information supplementation module 150 is used to supplement the key items according to the necessary information to generate complete order information.

[0063] The specific definition of the order completion system can be found in the definition of the order completion method above and will not be repeated here. Each module in the above-mentioned order completion system can be implemented in whole or in part through software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware format, or can be stored in the memory of the computer device in software format to facilitate the processor to call the corresponding operations of each of the above modules.

[0064] It should be noted that, in order to highlight the innovative part of the present invention, this embodiment does not introduce modules that are not closely related to solving the technical problem proposed by the present invention, but this does not mean that there are no other modules in this embodiment.

[0065] See Figure 3 The electronic device 1 may include a memory 12, a processor 13 and a bus, and may also include a computer program stored in the memory 12 and executable on the processor 13, such as an order completion program.

[0066] Among them, the memory 12 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 12 can be an internal storage unit of the electronic device 1, such as a mobile hard disk of the electronic device 1. In other embodiments, the memory 12 can also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Furthermore, the memory 12 can also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 12 can not only be used to store application software and various types of data installed on the electronic device 1, such as the code for completing an order, etc., but can also be used to temporarily store data that has been output or is to be output.

[0067] In some embodiments, the processor 13 may be composed of an integrated circuit, such as a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and a combination of various control chips. The processor 13 is the control core (Control Unit) of the electronic device 1. It utilizes various interfaces and circuits to connect the various components of the entire electronic device 1. It executes or runs programs or modules stored in the memory 12 (such as an order completion program) and calls data stored in the memory 12 to perform various functions of the electronic device 1 and process data.

[0068] The processor 13 executes the operating system and various installed applications of the electronic device 1. The processor 13 executes the applications to implement the steps in the above-mentioned order completion method.

[0069] Exemplarily, the computer program may be divided into one or more modules, which are stored in the memory 12 and executed by the processor 13 to complete the present application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution process of the computer program in the electronic device 1. For example, the computer program may be divided into a data acquisition order acquisition module 110, a type identification module 120, a key item detection module 130, an intent identification module 140, and an information supplementation module 150.

[0070] The above-mentioned integrated unit implemented in the form of a software function module can be stored in a computer-readable storage medium, which can be non-volatile or volatile. The above-mentioned software function module is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, computer device, or network device, etc.) or a processor to perform part of the functions of the order completion method described in various embodiments of the present application.

[0071] In summary, the present invention discloses an order completion method, system, device, and medium. Using keyword matching, the method accurately identifies the order type based on the order content and automatically detects whether all required key items in the order are complete. If the order is incomplete, the outbound call system is automatically triggered to initiate a dialogue with the user, allowing for faster access to the user's response information and alleviating information delays caused by manual follow-up. By inputting the user's response into a large language model for intent recognition, the necessary information provided by the user can be efficiently identified and extracted, thereby completing the missing key items and ensuring the completeness of the order information. This automated approach significantly improves the automation and accuracy of information collection, reduces manual intervention, lowers operating costs, and shortens processing time. It not only enhances the detail of order information, enabling drivers to make more accurate and rapid order acceptance decisions, but also improves overall order matching efficiency and satisfaction with completed orders. Furthermore, the present invention utilizes AI algorithms to achieve efficient and intelligent recommendations during the vehicle-cargo matching process, effectively reducing driver idle driving rates, optimizing transportation efficiency across the entire logistics chain, and increasing the platform's total transaction volume. By optimizing both information integrity and matching algorithms, this invention significantly improves order matching success rates and dispatch efficiency, reduces user wait times, and enhances driver job satisfaction. The large language model, with its strong contextual understanding and excellent conversational coherence, enables highly accurate recognition of user intent and precise extraction of cargo information during phone or text conversations. Therefore, this invention effectively overcomes the shortcomings of existing technologies and possesses high industrial value.

[0072] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical principles disclosed herein are intended to be covered by the claims of the present invention.

Claims

1. A method for completing an order, characterized in that: The method comprises: Get the user's order; Based on the keyword matching method, the order type is identified according to the content of the order; According to the order type, detecting key items of the order, and triggering an outbound call system when detecting that the key items are incomplete; wherein the outbound call system is used to establish a dialogue connection with the user; Receiving a user response returned by the outbound call system, and inputting the received user response into a large language model for intent recognition to identify necessary information; Complete the necessary information to the corresponding incomplete key items and generate complete order information.

2. The order completion method according to claim 1, characterized in that: The keyword matching method is used to identify the order type according to the content of the order, including: extracting text information from the order; Performing text cleaning and standardization processing on the text information to obtain pre-processed text information; Based on a preset word segmentation algorithm, the preprocessed text information is segmented to obtain a text word set; Extracting keywords from the text word set according to the keyword library; According to the type of the keyword, the corresponding order type is determined; wherein each type of keyword corresponds to one order type.

3. The order completion method according to claim 1, characterized in that: The receiving of the user response returned by the outbound call system and inputting the received user response into the large language model for intent recognition to identify necessary information therein include: Receive the user's reply, call the interface of the large language model, input the user's reply into the large language model, and determine whether the user's reply contains necessary information: If included, the necessary information is extracted; If not, a new guiding question is generated based on the user's reply, the new guiding question is fed back to the outbound calling system, and the outbound calling system is continuously monitored within a preset time period.

4. The order completion method according to claim 3, characterized in that: The process of generating the new guiding question includes: Extracting reply keywords of the user's reply; Determine whether the extracted answer keyword matches a preset question template; wherein each question in the question template corresponds to at least one answer keyword; If there is a match, the corresponding question is called as the new guiding question; If there is no match, the user response is input into a pre-trained dialogue generation language model to generate a new guiding question.

5. The order completion method according to claim 1, characterized in that: The necessary information is completed to the corresponding incomplete key items to generate complete order information, including: Inputting the necessary information into a pre-trained information type recognition model to identify the information type of the necessary information; wherein the information type recognition model is a natural language processing model; Based on the mapping relationship between the preset information type and the key items, the corresponding necessary information is filled in the preset position of each incomplete key item to generate complete order information.

6. The order completion method according to claim 1, characterized in that: After supplementing the key items according to necessary information and generating complete order information, the method further includes: matching drivers according to the order information based on a preset driver matching algorithm, determining drivers that match the order and making recommendations.

7. The order completion method according to claim 6, characterized in that: The driver matching algorithm based on the preset driver matching algorithm is used to match the driver according to the order information, and the driver matching the order is determined and recommended, including: Extracting key items from the order information; According to the key items, an initial set of drivers is selected from the driver database; Based on the preset multiple matching factors, the comprehensive matching degree of each initial driver in the initial driver set is calculated, and the driver with the highest comprehensive matching degree is screened out and recommended as the driver matching the order.

8. An order completion system, characterized in that: The system comprises: Order acquisition module, used to obtain user orders; A type identification module, used to identify the order type according to the content of the order based on a keyword matching method; A key item detection module, used to detect the key items of the order according to the order type, and trigger an outbound call system when it is detected that the key items are incomplete; wherein the outbound call system is used to contact the user; An intention recognition module is used to receive the user response returned by the outbound call system, and input the received user response into the large language model for intention recognition, thereby identifying necessary information therein; The information supplement module is used to supplement the key items according to necessary information to generate complete order information.

9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the order completion method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor of a computer, the computer is caused to execute the order completion method described in any one of claims 1 to 7.

Images