Method and device for optimizing medical consultation question-answering system, equipment and medium

Abstract

The invention relates to the field of medical treatment and health, and discloses a method, a device, equipment and a medium for optimizing a medical consultation question-answering system, in the method, questioning statements extracted in two modes not only can identify spontaneous high-frequency requirements of a user, but also can capture potential hotspots triggered by authoritative information, so that the user experience is improved. And the big model knowledge base of the medical consultation question-answering system is expanded by utilizing the preset question-answering statement generated by the question-answering statements extracted in the two modes, so that the question-answering accuracy and speed of the medical consultation question-answering system can be improved. The method comprises the steps of generating corresponding preset question statements based on a first question statement with a comprehensive score exceeding a preset comprehensive score and a second question statement with a comprehensive score exceeding the preset comprehensive score; inputting the preset question statement into a preset question and answer model to generate corresponding preset answer content; and storing the preset question statement and the corresponding preset answer content in a large model knowledge base in the medical consultation question and answer system.

Description

Technical Field

[0001] This invention relates to the fields of finance and healthcare, and in particular to a method, apparatus, equipment, and medium for optimizing a medical consultation and question-and-answer system. Background Technology

[0002] In the field of medical consultation, medical consultation question-and-answer systems, such as online medical platforms and intelligent consultation systems, can provide corresponding answers to users' questions. However, users often find the answers inaccurate and slow. Therefore, how to improve the accuracy and speed of medical consultation question-and-answer systems in answering questions has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0003] This invention provides a method, apparatus, equipment, and medium for optimizing a medical consultation question-and-answer system, in order to solve the technical problems of low accuracy and slow speed in answering questions in current medical consultation question-and-answer systems.

[0004] Firstly, a method for optimizing a medical consultation question-and-answer system is provided, including: Collect information on medical hot topics, and extract the first question statement from the information on medical hot topics; Collect the second question statement from the user's historical consultation data; Determine the overall scores corresponding to the first question statement and the second question statement respectively; Determine whether the current time is an off-peak time; if it is an off-peak time, generate corresponding preset question statements based on the first question statement whose comprehensive score exceeds the preset comprehensive score and the second question statement whose comprehensive score exceeds the preset comprehensive score, respectively. The preset question statement is input into the preset question-and-answer model to generate the corresponding preset answer content; The preset question statements and corresponding preset answer contents are stored in the large model knowledge base of the medical consultation question and answer system.

[0005] In some embodiments, the step of determining the overall score corresponding to the second question statement includes: Based on the second question, determine the user's behavioral trend score; Determine whether the behavioral trend score exceeds a preset behavioral trend score; If the score exceeds the preset behavioral trend score, then the correlation score between the second question and the medical hot topic information is determined. Based on the user's behavioral trend score and the relevance score, a comprehensive score corresponding to the second question statement is calculated.

[0006] In some embodiments, the step of determining the user's behavioral trend score based on the second question statement includes: The time and number of times medical-related consultation topics were asked in the second question statement within a preset time period were statistically analyzed. Obtain the amount paid by the user within the preset time period; A time interval score is determined based on the time interval between the user's last question about the consultation topic within a preset time and the current time. Based on the number of queries, determine the query count score; Based on the payment amount, a payment amount score is determined; The user's behavioral trend score is determined based on the time interval score, the number of inquiries score, and the amount paid score.

[0007] In some embodiments, the step of determining the relevance score between the second question statement and the medical hot topic information includes: Identify the hot topics within the aforementioned medical hot topics information; Determine the relevance score between the consultation topic and the hot topic.

[0008] In some embodiments, the step of calculating the comprehensive score corresponding to the second question statement based on the user's behavioral trend score and the relevance score includes: Calculate the product of the user's behavioral trend score and the correlation score; Divide the product by the preset total score to obtain the comprehensive score corresponding to the second question statement.

[0009] In some embodiments, the method further includes: Received the third question input from the user; Search the large model knowledge base for the preset question statement that has the highest similarity to the third question statement; Determine and output the preset answer content corresponding to the preset question statement with the highest similarity to the third question statement.

[0010] In some embodiments, the large model knowledge base includes the VikingDB vector database; the step of finding the preset question statement with the highest similarity to the third question statement from the large model knowledge base includes: Based on the third question statement, determine the retrieval score and rearrangement score corresponding to the preset question statement in the VikingDB vector database; Filter out preset question statements whose search score is greater than a first preset search score and whose rearrangement score is greater than a first preset rearrangement score, preset question statements whose search score is greater than a second preset search score, or preset question statements whose rearrangement score is greater than a second preset rearrangement score; The first preset question among the selected preset question statements is determined as the preset question statement with the highest similarity to the third question statement.

[0011] Secondly, an apparatus for optimizing a medical consultation and question-and-answer system is provided, comprising: The first acquisition unit is used to acquire medical hotspot information and extract the first question statement from the medical hotspot information. The second data collection unit is used to collect the second question statement from the user's historical consultation data; The first determining unit is used to determine the comprehensive scores corresponding to the first question statement and the second question statement, respectively. The first judgment unit is used to determine whether the current time is an idle time. The first generation unit is used to generate corresponding preset question statements based on the first question statement and the second question statement whose comprehensive score exceeds the preset comprehensive score, respectively, if it is an idle time. The second generation unit is used to input the preset question statement into the preset question-answering model to generate the corresponding preset answer content; The storage unit is used to store the preset question statements and corresponding preset answer content into the large model knowledge base of the medical consultation question and answer system.

[0012] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method for optimizing a medical consultation question-and-answer system as described above.

[0013] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method for optimizing a medical consultation question-and-answer system as described above.

[0014] The above-mentioned optimized medical consultation question-and-answer system, device, equipment, and medium can collect medical hot topic information and extract a first question statement from the medical hot topic information; collect a second question statement from the user's historical consultation data; determine the comprehensive scores corresponding to the first question statement and the second question statement respectively; determine whether the current time is off-peak; if it is off-peak, generate corresponding preset question statements based on the first question statement and the second question statement whose comprehensive scores exceed the preset comprehensive scores respectively; input the preset question statements into the preset question-and-answer model to generate corresponding preset answer content; and store the preset question statements and the corresponding preset answer content in the large model knowledge base of the medical consultation question-and-answer system. In this invention, a first question statement is collected through medical hot topic information, and a second question statement is collected through historical consultation data. Corresponding preset question statements are generated based on the first and second question statements, respectively, whose comprehensive scores exceed preset comprehensive scores. This method of extracting question statements not only identifies high-frequency spontaneous user needs but also captures potential hot topics triggered by authoritative information. Furthermore, the preset question statements generated from these two methods are used to expand the large model knowledge base of the medical consultation question-and-answer system, improving the accuracy and speed of the system's responses. Additionally, expanding the large model knowledge base during off-peak hours avoids resource shortages during peak periods and improves server load balancing. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of an application environment for a method of optimizing a medical consultation question-and-answer system according to an embodiment of the present invention; Figure 2 A flowchart illustrating a method for optimizing a medical consultation question-and-answer system provided in an embodiment of the present invention; Figure 3 An exemplary schematic diagram of an apparatus for optimizing a medical consultation question-and-answer system is shown according to some embodiments; Figure 4 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 5 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] In the field of medical consultation, medical consultation question-and-answer systems, such as online medical platforms and intelligent consultation systems, can provide corresponding answers to users' questions. However, users often find the answers inaccurate and slow. Therefore, how to improve the accuracy and speed of answers provided by medical consultation question-and-answer systems has become a pressing technical problem for those skilled in the art.

[0019] The method for optimizing a medical consultation question-and-answer system provided in this embodiment of the invention can be applied to, for example... Figure 1 In this application environment, the user inputs a second question through the client, and the medical answer model outputs the corresponding answer. The server collects medical hotspot information and extracts a first question from it; it also collects a second question from the user's historical consultation data; it determines the comprehensive score corresponding to the first and second questions; it checks if it is currently a quiet period; if so, it generates corresponding preset questions based on the first and second questions whose comprehensive scores exceed a preset score; it inputs these preset questions into a preset question-and-answer model to generate corresponding preset answer content; and it stores the preset questions and their corresponding preset answer content in the large model knowledge base of the medical consultation question-and-answer system. In this invention, a first question statement is collected through medical hot topic information, and a second question statement is collected through historical consultation data. Corresponding preset question statements are generated based on the first and second question statements, respectively, whose comprehensive scores exceed preset comprehensive scores. This method of extracting question statements not only identifies high-frequency spontaneous user needs but also captures potential hot topics triggered by authoritative information. Furthermore, the preset question statements generated from these two methods are used to expand the large model knowledge base of the medical consultation question-and-answer system, improving the accuracy and speed of the system's responses. Additionally, expanding the large model knowledge base during off-peak hours avoids resource shortages during peak periods and improves server load balancing.

[0020] The user terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster consisting of multiple servers. The invention will now be described in detail through specific embodiments.

[0021] Please see Figure 2 As shown, Figure 2 A flowchart illustrating the method for optimizing medical consultation Q&A provided in an embodiment of the present invention includes the following steps S100-S700.

[0022] S100. Collect medical hotspot information and extract the first question statement from the medical hotspot information.

[0023] In this embodiment of the application, medical hot topics information can be collected not only from news, but also by manually sorting other publicly available data. For example, medical hot topics frequently discussed by the public can be identified from short video data.

[0024] In this embodiment, a first question and its corresponding answer can be extracted from medical hot topic information. In one example, an interview dialogue similar to that between a host and a doctor is found from popular science articles on medical hot topics. The host's question is extracted as the first question, and the doctor's answer is extracted as the corresponding answer.

[0025] Thus, in addition to relying on users' historical consultation data, this application embodiment can also conduct in-depth mining and integration of external data such as widely publicized medical news, i.e., medical hot topic information, and can capture potential medical hot topics in advance through medical hot topic information.

[0026] S200: Collect the second question statement from the user's historical consultation data.

[0027] In this embodiment, the process of generating historical consultation data includes the user inputting a second question into a large-scale medical answer model, which then outputs the answer content corresponding to the second question. The second question and its corresponding answer content are stored as historical consultation data. The large-scale medical answer model can be an existing large-scale model of medical information, through which users can ask questions and receive corresponding answers.

[0028] In this embodiment of the application, the second question statement is collected from the user's historical consultation data, which means extracting the second question statement from the stored historical consultation data.

[0029] It should be noted that the user's historical consultation data was obtained with the user's consent.

[0030] S300. Determine the comprehensive scores corresponding to the first question statement and the second question statement respectively.

[0031] Due to limited storage space and other issues, the large model knowledge base mentioned later in this application embodiment is expanded to focus on medical content that is currently of greatest concern to the public, namely, medical hot topics. Therefore, it is necessary to determine the relevance between question statements and medical hot topics so that subsequent question statements with high relevance to medical hot topics can be expanded and corresponding answer content generated and stored in the large model knowledge base. This allows for the storage of dialogue content with high relevance to medical hot topics (i.e., the preset question statements and corresponding preset answer content mentioned later) within a limited storage space.

[0032] In this embodiment, a comprehensive score is used to represent the relevance of the question to trending medical information. The higher the comprehensive score, the stronger the relevance between the question and the trending medical information.

[0033] In some embodiments, the step of determining the overall score corresponding to the second question statement includes: Based on the second question, the user's behavioral trend score is determined.

[0034] In this embodiment, the user's behavioral trend score can describe the user's activity level, payment history, and whether the questions asked are related to healthcare, etc. Due to storage space limitations in the large model knowledge base, users with high behavioral trend scores need to be given special attention. Specifically, the questions that users with high behavioral trend scores might ask and the corresponding answers should be pre-stored in the large model knowledge base to improve the speed and accuracy of obtaining corresponding answers when these users subsequently ask questions using the large model knowledge base.

[0035] In some embodiments, the step of determining the user's behavioral trend score based on the second question statement includes: The time and number of inquiries related to medical issues in the second question statement within a preset time period are statistically analyzed.

[0036] In one example, the preset timeframe is one week. The first topic related to health could include colds, medication, and drug names.

[0037] Obtain the amount paid by the user within the preset time period.

[0038] In this embodiment of the application, during the historical consultation process, the user may pay for medical consultation or pay for medicine. The total amount of all paid items within a preset time period is calculated to obtain the total amount paid within the preset time period.

[0039] A time interval score is determined based on the time interval between the user's last question on a medical-related consultation topic within a preset time period and the current time.

[0040] In this embodiment of the application, when a user asks a question using the medical answer model, there is a corresponding question time. The time interval between the question time and the current time is calculated, and a time interval score is determined based on this time interval.

[0041] In one example, the maximum score for the time interval is 5 points. A time interval of 1 day or less is assigned a score of 5. A time interval greater than 1 day and less than or equal to 2 days is assigned a score of 4. A time interval greater than 2 days and less than or equal to 3 days is assigned a score of 3. A time interval greater than 3 days and less than or equal to 4 days is assigned a score of 2. A time interval greater than 2 days and less than or equal to 5 days is assigned a score of 1. A time interval greater than 1 day and less than or equal to 6 days is assigned a score of 0.

[0042] Based on the number of queries, determine the query count score; In one example, the maximum score for the number of queries is 5 points. A score of 5 points is awarded for more than 40 queries. A score of 4 points is awarded for 30 to 40 queries. A score of 3 points is awarded for 20 to 30 queries. A score of 2 points is awarded for 10 to 20 queries. A score of 1 point is awarded for 5 to 10 queries. A score of 0 points is awarded for 5 to 5 queries.

[0043] Based on the payment amount, a payment amount score is determined; In one example, the score for the payment amount is 5 points. A payment amount greater than 1000 yuan receives 5 points. A payment amount between 1000 yuan and 600 yuan receives 4 points. A payment amount between 600 yuan and 400 yuan receives 3 points. A payment amount between 400 yuan and 200 yuan receives 2 points. A payment amount between 200 yuan and 0 yuan receives 1 point. A payment amount of 0 yuan receives 0 points.

[0044] The user's behavioral trend score is determined based on the time interval score, the number of inquiries score, and the amount paid score.

[0045] In this embodiment of the application, the sum of the time interval score, the number of inquiries score, and the payment amount score is determined as the user's behavior trend score. Specifically, the behavior trend score U = time interval score r + number of inquiries score f + payment amount score m.

[0046] Determine whether the behavioral trend score exceeds a preset behavioral trend score.

[0047] In this embodiment, if a user's behavioral trend score exceeds a preset behavioral trend score, it indicates that the user may be relatively active, their questions are related to current trending topics, and / or they are willing to pay for services. In one example, the preset behavioral trend score is 8 points.

[0048] If the score exceeds the preset behavioral trend score, then the correlation score between the second question statement and the medical hot topic information is determined.

[0049] In this embodiment, when a user's behavioral trend score exceeds a preset behavioral trend score, it indicates that the user needs to be given special attention. Furthermore, since the large model knowledge base is primarily used in the medical field to provide users with more accurate medical information, it is necessary to determine the relevance between the second question and the medical hot topic information, specifically represented by a relevance score. A high relevance score indicates a strong correlation between the second question and the medical hot topic information. A low relevance score indicates a weak correlation between the second question and the medical hot topic information.

[0050] In some embodiments, the step of determining the relevance score between the second question statement and the medical hot topic information includes: Identify the hot topics within the aforementioned medical hot topics information.

[0051] In this embodiment, the hot topics can be those mentioned more frequently than a preset frequency in medical hot topics information. In this embodiment, the preset frequency can be set according to actual needs and is not limited here.

[0052] Determine the relevance score between medical-related consultation topics and the aforementioned hot topics.

[0053] In this embodiment, a target large model is trained using a training set. The training set includes multiple training data points, such as preset medical-related consultation topics, preset hot topics, and corresponding preset correlation scores. In this embodiment, the medical-related consultation topics and hot topics are input into the trained target large model to obtain the corresponding correlation scores. For example, a hot topic could be influenza. A medical-related consultation topic could be the common cold; in this case, the correlation scores between influenza and the common cold are determined.

[0054] Based on the user's behavioral trend score and the relevance score, a comprehensive score corresponding to the second question statement is calculated.

[0055] In some embodiments, the step of calculating the comprehensive score corresponding to the second question statement based on the user's behavior trend score and the relevance score includes: calculating the product of the user's behavior trend score and the relevance score; and dividing the product by a preset total score to obtain the comprehensive score corresponding to the second question statement.

[0056] In this embodiment of the application, the preset total score is the product of the user's behavior trend score and correlation score. For example, if the user's behavior trend score is 15 points and the correlation score is 10 points, the preset total score is 150.

[0057] In one example, the user's behavioral trend score is 9 points, the relevance score is 10 points, and the preset total score is 150. Therefore, the comprehensive score corresponding to the second question statement is determined to be 90 / 150, which is 0.6.

[0058] In this embodiment of the application, the comprehensive score corresponding to the first question statement is determined as the target comprehensive score. For example, the target comprehensive score can be 1. In this embodiment of the application, since the first question statement is extracted from medical hot topic information, the target comprehensive score corresponding to the first question statement is set to be greater than the preset comprehensive score.

[0059] S400: Determine if the current time is idle.

[0060] In this embodiment, computing resources (such as GPU computing power, NLP model inference threads, cache storage space, etc.) are idle during off-peak hours (i.e., off-peak hours, such as early morning or weekday days when consultations are not peak), while they face the risk of overload during peak hours. This embodiment utilizes knowledge of off-peak resource preprocessing to avoid resource contention during peak hours.

[0061] S500. If it is an off-peak time, then generate corresponding preset question statements based on the first question statement whose comprehensive score exceeds the preset comprehensive score and the second question statement whose comprehensive score exceeds the preset comprehensive score.

[0062] In this embodiment of the application, if the overall score corresponding to the second question exceeds a preset overall score, it indicates that the user needs to be given special attention, and the second question asked by the user who needs special attention is about a popular topic. In one example, the preset overall score is 0.8.

[0063] In this embodiment, the first question statement with a comprehensive score exceeding the preset comprehensive score and the second question statement with a comprehensive score exceeding the preset comprehensive score are both question statements related to medical hot topics. Based on the question statements related to medical hot topics, the large model knowledge base in the following text is expanded, which can increase the probability that the large model knowledge base hits the question statement raised by the user in real time (i.e., the third question statement), and thus quickly and accurately determine the answer content corresponding to the question statement raised by the user in real time from the large model knowledge base.

[0064] In one example, the idle time is from 0:00 to 6:00. Expanding the large model knowledge base can utilize the idle CPU / GPU and distributed storage bandwidth from 0:00 to 6:00.

[0065] In this embodiment, existing AI models can be used to generate highly similar preset question statements based on the first and second question statements whose comprehensive scores exceed a preset comprehensive score. For example, the second question statement could be "How to treat a mosquito bite," and the corresponding preset question statement could be something like "The difference between a common mosquito bite and chikungunya fever."

[0066] S600. Input the preset question statement into the preset question-answering model to generate the corresponding preset answer content.

[0067] In this embodiment, the preset question-and-answer model can be an existing large-scale AI model. The existing large-scale AI model generates corresponding preset answer content based on preset question statements.

[0068] S700. Store the preset question statement and the corresponding preset answer content into the large model knowledge base of the medical consultation question and answer system.

[0069] In this way, based on the first and second question statements whose comprehensive scores exceed the preset comprehensive scores, corresponding preset question statements are generated, and then corresponding preset answer content is generated and stored in the large model knowledge base of the medical consultation question and answer system. This can increase the probability of the large model knowledge base hitting the question statement (i.e. the third question statement) raised by the user in real time, and improve the accuracy and speed of obtaining the corresponding answer content.

[0070] In this embodiment of the application, when a user actually uses the medical consultation question and answer system, after entering a question, the corresponding answer content is retrieved from the large model knowledge base of the medical consultation question and answer system and used as the output content of the medical consultation question and answer system.

[0071] In some embodiments, the method further includes: periodically extracting a preset question statement and its corresponding preset answer content from the large model knowledge base; manually verifying the extracted preset question statement and its corresponding preset answer content; and deleting preset question statements and their corresponding preset answer content that fail verification.

[0072] In this embodiment of the application, in order to improve the prediction accuracy of the medical consultation question and answer system, manual intervention is performed on the large model knowledge base in advance, and the preset question statements and corresponding preset answer contents that fail to be verified in the large model knowledge base are deleted.

[0073] In some embodiments, the method further includes: Received the third question input from the user; Search the large model knowledge base for the preset question statement that has the highest similarity to the third question statement; In some embodiments, the large model knowledge base includes the VikingDB vector database; the step of finding the preset question statement with the highest similarity to the third question statement from the large model knowledge base includes: Based on the third question statement, determine the retrieval score and rearrangement score corresponding to the preset question statement in the VikingDB vector database; Filter out preset question statements whose search score is greater than a first preset search score and whose rearrangement score is greater than a first preset rearrangement score, preset question statements whose search score is greater than a second preset search score, or preset question statements whose rearrangement score is greater than a second preset rearrangement score; The first preset question among the selected preset question statements is determined as the preset question statement with the highest similarity to the third question statement.

[0074] Determine and output the preset answer content corresponding to the preset question statement with the highest similarity to the third question statement.

[0075] The method in this application embodiment can shorten the response time from the traditional 300-500ms to less than 100ms (the measured average response time is 80ms). Users do not need to wait for complex semantic analysis or database retrieval, which is especially suitable for quick answers to urgent medical questions (such as "What medicine should I give my child if he has a fever"), and significantly improves user satisfaction.

[0076] As can be seen, in the above solution, the first question is collected through medical hot topic information, and the second question is collected through historical consultation data. Corresponding preset question statements are generated based on the first and second question statements, respectively, whose comprehensive scores exceed preset comprehensive scores. This method of extracting question statements not only identifies high-frequency spontaneous user needs but also captures potential hot topics triggered by authoritative information. Furthermore, the preset question statements generated from these two methods are used to expand the large model knowledge base of the medical consultation question-and-answer system, improving the accuracy and speed of the system's responses. In addition, expanding the large model knowledge base during off-peak hours avoids resource shortages during peak periods, while simultaneously improving server load balancing by 40% and reducing overall operating costs.

[0077] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0078] In one embodiment, an apparatus for optimizing a medical consultation question-and-answer system is provided, which corresponds one-to-one with the method for optimizing a medical consultation question-and-answer system described in the above embodiments. For example... Figure 3 As shown, the device for optimizing the medical consultation question-and-answer system includes: a first acquisition unit 301, a second acquisition unit 302, a first determination unit 303, a first judgment unit 304, a first generation unit 305, a second generation unit 306, and a storage unit 307. Detailed descriptions of each functional module are as follows: The first acquisition unit 301 is used to acquire medical hotspot information and extract the first question statement from the medical hotspot information; The second acquisition unit 302 is used to acquire the second question statement from the user's historical consultation data; The first determining unit 303 is used to determine the comprehensive scores corresponding to the first question statement and the second question statement, respectively. The first judgment unit 304 is used to determine whether the current time is an idle time; The first generation unit 305 is used to generate corresponding preset question statements based on the first question statement and the second question statement whose comprehensive score exceeds the preset comprehensive score, respectively, if it is an idle time. The second generation unit 306 is used to input the preset question statement into the preset question-answering model to generate the corresponding preset answer content; Storage unit 307 is used to store the preset question statement and the corresponding preset answer content into the large model knowledge base of the medical consultation question and answer system.

[0079] In some embodiments, the first determining unit includes: The second determining unit is used to determine the user's behavioral trend score based on the second question statement; The second judgment unit is used to determine whether the behavior trend score exceeds the preset behavior trend score. The third determining unit is used to determine the correlation score between the second question statement and the medical hot topic information if the score exceeds the preset behavioral trend score. The calculation unit is used to calculate the comprehensive score corresponding to the second question statement based on the user's behavior trend score and the relevance score.

[0080] In some embodiments, the second determining unit includes: The statistics unit is used to count the time and number of inquiries about medical-related consultation topics in the second question statement within a preset time period; The acquisition unit is used to acquire the amount paid by the user within the preset time period; The fourth determining unit is used to determine the time interval score based on the time interval between the user's last question about a medical-related consultation topic within a preset time and the current time. The fifth determining unit is used to determine the number of queries score based on the number of queries. The sixth determining unit is used to determine the payment amount fraction based on the payment amount; The seventh determining unit is used to determine the user's behavioral trend score based on the time interval score, the number of inquiries score, and the payment amount score.

[0081] In some embodiments, the third determining unit includes: The eighth determining unit is used to determine the hot topics in the medical hot topic information; The ninth determining unit is used to determine the relevance score between medical-related consultation topics and the aforementioned hot topics.

[0082] In some embodiments, the computing unit is specifically used for: Calculate the product of the user's behavioral trend score and the correlation score; Divide the product by the preset total score to obtain the comprehensive score corresponding to the second question statement.

[0083] In some embodiments, the apparatus further includes: The receiving unit is used to receive the third query statement input by the user; The search unit is used to search for the preset question statement that has the highest similarity to the third question statement from the large model knowledge base; The tenth determining unit is used to determine and output the preset answer content corresponding to the preset question statement with the highest similarity to the third question statement.

[0084] In some embodiments, the large model knowledge base includes the VikingDB vector database; the search unit is specifically used for: Based on the third question statement, determine the retrieval score and rearrangement score corresponding to the preset question statement in the VikingDB vector database; Filter out preset question statements whose search score is greater than a first preset search score and whose rearrangement score is greater than a first preset rearrangement score, preset question statements whose search score is greater than a second preset search score, or preset question statements whose rearrangement score is greater than a second preset rearrangement score; The first preset question among the selected preset question statements is determined as the preset question statement with the highest similarity to the third question statement.

[0085] This invention provides a device for optimizing a medical consultation question-and-answer system. It collects a first question statement from medical hot topic information and a second question statement from historical consultation data. Based on the first and second question statements, which each have a composite score exceeding a preset score, corresponding preset question statements are generated. This method of extracting question statements not only identifies frequently occurring user needs but also captures potential hot topics triggered by authoritative information. Furthermore, the preset question statements generated from these two methods are used to expand the large-scale knowledge base of the medical consultation question-and-answer system, improving the accuracy and speed of answering questions. Additionally, expanding the large-scale knowledge base during off-peak hours avoids resource shortages during peak periods, while simultaneously improving server load balancing by 40% and reducing overall operating costs.

[0086] Specific limitations regarding the device for optimizing the medical consultation question-and-answer system can be found in the limitations regarding the method for optimizing the medical consultation question-and-answer system described above, and will not be repeated here. Each module in the aforementioned device for optimizing the medical consultation question-and-answer system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0087] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the server-side functions or steps of a method for optimizing a medical consultation question-and-answer system.

[0088] In one embodiment, a computer device is provided, which may be a user terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the user-side functions or steps of a method for optimizing a medical consultation question-and-answer system.

[0089] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Collect information on medical hot topics, and extract the first question statement from the information on medical hot topics; Collect the second question statement from the user's historical consultation data; Determine the overall scores corresponding to the first question statement and the second question statement respectively; Determine whether the current time is an off-peak time; if it is an off-peak time, generate corresponding preset question statements based on the first question statement whose comprehensive score exceeds the preset comprehensive score and the second question statement whose comprehensive score exceeds the preset comprehensive score, respectively. The preset question statement is input into the preset question-and-answer model to generate the corresponding preset answer content; The preset question statements and corresponding preset answer contents are stored in the large model knowledge base of the medical consultation question and answer system.

[0090] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Collect information on medical hot topics, and extract the first question statement from the information on medical hot topics; Collect the second question statement from the user's historical consultation data; Determine the overall scores corresponding to the first question statement and the second question statement respectively; Determine whether the current time is an off-peak time; if it is an off-peak time, generate corresponding preset question statements based on the first question statement whose comprehensive score exceeds the preset comprehensive score and the second question statement whose comprehensive score exceeds the preset comprehensive score, respectively. The preset question statement is input into the preset question-and-answer model to generate the corresponding preset answer content; The preset question statements and corresponding preset answer contents are stored in the large model knowledge base of the medical consultation question and answer system.

[0091] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0092] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0093] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0094] The software tools or components not belonging to our company that appear in the embodiments of this application are merely examples and do not represent actual use.

[0095] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for optimizing a medical consultation question-and-answer system, characterized in that, include: Collect information on medical hot topics, and extract the first question statement from the information on medical hot topics; Collect the second question statement from the user's historical consultation data; Determine the overall scores corresponding to the first question statement and the second question statement respectively; Determine if the current time is an idle period; If it is during off-peak hours, then based on the first question statement whose comprehensive score exceeds the preset comprehensive score and the second question statement whose comprehensive score exceeds the preset comprehensive score, corresponding preset question statements are generated respectively; The preset question statement is input into the preset question-and-answer model to generate the corresponding preset answer content; The preset question statements and corresponding preset answer contents are stored in the large model knowledge base of the medical consultation question and answer system.

2. The method according to claim 1, characterized in that, The steps to determine the overall score corresponding to the second question statement include: Based on the second question, determine the user's behavioral trend score; Determine whether the behavioral trend score exceeds a preset behavioral trend score; If the score exceeds the preset behavioral trend score, then the correlation score between the second question and the medical hot topic information is determined. Based on the user's behavioral trend score and the relevance score, a comprehensive score corresponding to the second question statement is calculated.

3. The method according to claim 2, characterized in that, The step of determining the user's behavioral trend score based on the second question includes: The time and number of times medical-related consultation topics were asked in the second question statement within a preset time period were statistically analyzed. Obtain the amount paid by the user within the preset time period; The time interval score is determined based on the time interval between the user's last question on a medical-related consultation topic within a preset time and the current time. Based on the number of queries, determine the query count score; Based on the payment amount, a payment amount score is determined; The user's behavioral trend score is determined based on the time interval score, the number of inquiries score, and the amount paid score.

4. The method according to claim 3, characterized in that, The step of determining the relevance score between the second question and the medical hot topic information includes: Identify the hot topics within the aforementioned medical hot topics information; Determine the relevance score between medical-related consultation topics and the aforementioned hot topics.

5. The method according to claim 1, characterized in that, The step of calculating the comprehensive score corresponding to the second question statement based on the user's behavioral trend score and the relevance score includes: Calculate the product of the user's behavioral trend score and the correlation score; Divide the product by the preset total score to obtain the comprehensive score corresponding to the second question statement.

6. The method according to claim 1, characterized in that, Also includes: Received the third question input from the user; Search the large model knowledge base for the preset question statement that has the highest similarity to the third question statement; Determine and output the preset answer content corresponding to the preset question statement with the highest similarity to the third question statement.

7. The method according to claim 6, characterized in that, The large model knowledge base includes the VikingDB vector database; the step of finding the preset question statement with the highest similarity to the third question statement from the large model knowledge base includes: Based on the third question statement, determine the retrieval score and rearrangement score corresponding to the preset question statement in the VikingDB vector database; Filter out preset question statements whose search score is greater than a first preset search score and whose rearrangement score is greater than a first preset rearrangement score, preset question statements whose search score is greater than a second preset search score, or preset question statements whose rearrangement score is greater than a second preset rearrangement score; The first preset question among the selected preset question statements is determined as the preset question statement with the highest similarity to the third question statement.

8. A device for optimizing a medical consultation and question-and-answer system, characterized in that, include: The first acquisition unit is used to acquire medical hotspot information and extract the first question statement from the medical hotspot information. The second data collection unit is used to collect the second question statement from the user's historical consultation data; The first determining unit is used to determine the comprehensive scores corresponding to the first question statement and the second question statement, respectively. The first judgment unit is used to determine whether the current time is an idle time. The first generation unit is used to generate corresponding preset question statements based on the first question statement and the second question statement whose comprehensive score exceeds the preset comprehensive score, respectively, if it is an idle time. The second generation unit is used to input the preset question statement into the preset question-answering model to generate the corresponding preset answer content; The storage unit is used to store the preset question statements and corresponding preset answer content into the large model knowledge base of the medical consultation question and answer system.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for optimizing a medical consultation question-and-answer system as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method for optimizing a medical consultation question-and-answer system as described in any one of claims 1 to 7.

Images