Information retrieval method and device and storage medium

Abstract

The invention discloses an information retrieval method and device and a storage medium, relates to the technical field of communication, and is used for improving the accuracy of information retrieval. The method comprises the following steps: acquiring a plurality of sample knowledge points; according to the plurality of sample knowledge points and a large language model, obtaining semantics of each sample knowledge point; according to the semantics of each sample knowledge point, determining the relationship between any two sample knowledge points; the relationship between any two sample knowledge points is used for retrieving the sample knowledge points. The method and the device are applied to an information retrieval process.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to an information retrieval method, device and storage medium. Background Art

[0002] In related technologies, large language models are based on deep learning technology, which requires a large number of samples and computing power, and a long period of training to achieve the ideal intelligent question-answering effect. When samples and computing power are insufficient, the Retrieval Augmented Generation (RAG) model is usually used to index and store knowledge. When a question is received, the question is answered with the help of the trained large language model.

[0003] Although the RAG model can answer questions from the knowledge base, due to the simple way of storing and retrieving knowledge points in the knowledge base, the RAG model cannot deeply understand the content of the knowledge points, resulting in low accuracy of information retrieval. Therefore, how to improve the accuracy of information retrieval is a technical problem that still needs to be solved. Summary of the invention

[0004] The present application provides an information retrieval method, device and storage medium for improving the accuracy of information retrieval by a user terminal.

[0005] In order to achieve the above objectives, this application adopts the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides an information retrieval method, wherein an information retrieval device obtains a plurality of sample knowledge points; the information retrieval device obtains the semantics of each sample knowledge point based on the plurality of sample knowledge points and a large language model; the information retrieval device determines the relationship between any two sample knowledge points based on the semantics of each of the sample knowledge points; the relationship between any two sample knowledge points is used to retrieve the sample knowledge points.

[0007] In a possible implementation manner, the sample knowledge point and the relationship between any two sample knowledge points are stored in a vector library.

[0008] In a possible implementation, retrieval request information is received from a terminal device; the retrieval request information includes at least one keyword; based on the at least one keyword and the large language model, a first sample knowledge point corresponding to the at least one keyword is determined; the first sample knowledge point is a sample knowledge point among the multiple sample knowledge points; a first relationship is obtained from the relationship between any two sample knowledge points; the first relationship is a relationship between the first sample knowledge point and a knowledge point among the multiple sample knowledge points other than the first sample knowledge point; based on the at least one keyword, the first sample knowledge point and the first relationship, a second sample knowledge point is determined; and based on the first sample knowledge point and the second sample knowledge point, retrieval response information is generated.

[0009] In a possible implementation, a key sentence is generated based on the at least one keyword; and the second sample knowledge point is determined based on the key sentence, the first sample knowledge point and the first relationship.

[0010] In one possible implementation, determine whether the first sample knowledge point and the second sample knowledge point satisfy the condition for generating the retrieval response information; when the first sample knowledge point and the second sample knowledge point do not satisfy the condition for generating the retrieval response information, obtain the second relationship of the second sample knowledge point; the second relationship is the relationship between the second sample knowledge point and the knowledge points among the multiple sample knowledge points except the second sample knowledge point; determine the third sample knowledge point based on the at least one keyword, the second sample knowledge point and the second relationship; generate the retrieval response information based on the first sample knowledge point, the second sample knowledge point and the third sample knowledge point.

[0011] In a possible implementation, the multiple sample knowledge points are input into the large language model to obtain prediction results; the prediction results are used to characterize the semantics of each sample knowledge point; when the content volume of the prediction results is less than a threshold, in response to a sample marking operation on each of the sample knowledge points, the semantics corresponding to the marking operation is determined as the semantics of the sample knowledge point.

[0012] In the second aspect, the present application provides an information retrieval device, including: an acquisition unit and a processing unit; the acquisition unit is used to acquire multiple sample knowledge points; the acquisition unit is also used to acquire the semantics of each of the sample knowledge points based on the multiple sample knowledge points and the large language model; the processing unit is used to determine the relationship between any two sample knowledge points based on the semantics of each of the sample knowledge points; the relationship between any two sample knowledge points is used to retrieve the sample knowledge points.

[0013] In a possible implementation, the processing unit is further configured to store the sample knowledge point and the relationship between any two sample knowledge points in a vector library.

[0014] In a possible implementation, a communication unit is used to receive retrieval request information from a terminal device; the retrieval request information includes at least one keyword; the processing unit is also used to determine a first sample knowledge point corresponding to the at least one keyword based on the at least one keyword and the large language model; the first sample knowledge point is a sample knowledge point among the multiple sample knowledge points; the processing unit is also used to obtain a first relationship from the relationship between any two sample knowledge points; the first relationship is a relationship between the first sample knowledge point and a knowledge point among the multiple sample knowledge points other than the first sample knowledge point; the processing unit is also used to determine a second sample knowledge point based on the at least one keyword, the first sample knowledge point and the first relationship; the processing unit is also used to generate retrieval response information based on the first sample knowledge point and the second sample knowledge point.

[0015] In a possible implementation, the processing unit is further used to generate a key sentence based on the at least one keyword; the processing unit is further used to determine the second sample knowledge point based on the key sentence, the first sample knowledge point and the first relationship.

[0016] In one possible implementation, the processing unit is further used to determine whether the first sample knowledge point and the second sample knowledge point satisfy the condition for generating the retrieval response information; the processing unit is further used to instruct the acquisition unit to acquire the second relationship of the second sample knowledge point when the first sample knowledge point and the second sample knowledge point do not satisfy the condition for generating the retrieval response information; the second relationship is the relationship between the second sample knowledge point and the knowledge points among the multiple sample knowledge points except the second sample knowledge point; the processing unit is further used to determine the third sample knowledge point based on the at least one keyword, the second sample knowledge point and the second relationship; the processing unit is further used to generate the retrieval response information based on the first sample knowledge point, the second sample knowledge point and the third sample knowledge point.

[0017] In a possible implementation, the processing unit is further used to input the multiple sample knowledge points into the large language model to obtain a prediction result; the prediction result is used to characterize the semantics of each sample knowledge point; the processing unit is further used to, in a case where the content volume of the prediction result is less than a threshold, respond to a sample marking operation on each of the sample knowledge points and determine the semantics corresponding to the marking operation as the semantics of the sample knowledge point.

[0018] In a third aspect, the present application provides an information retrieval device, which includes: a processor and a memory; wherein the memory is used to store computer execution instructions, and when the information retrieval device is running, the processor executes the computer execution instructions stored in the memory to enable the information retrieval device to perform the information retrieval method described in the first aspect and any possible implementation of the first aspect.

[0019] In a fourth aspect, the present application provides a computer-readable storage medium, which stores instructions. When the instructions in the computer-readable storage medium are executed by a processor of an information retrieval device, the information retrieval device can perform the information retrieval method described in the first aspect and any possible implementation of the first aspect.

[0020] In a fifth aspect, an embodiment of the present application provides a chip, the chip including a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instructions to implement the information retrieval method described in the above-mentioned first aspect and any possible implementation method of the first aspect.

[0021] In a sixth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the information retrieval method described in the first aspect and any possible implementation manner of the first aspect.

[0022] These and other aspects of the present application will become more apparent from the following description.

[0023] The above scheme brings at least the following beneficial effects: In the embodiment of the present application, the information retrieval device obtains multiple sample knowledge points, and obtains the semantics of each sample knowledge point in the multiple sample knowledge points based on the large language model. In this way, the information retrieval device can establish a relationship between any two sample knowledge points based on the semantics of each sample knowledge point in the multiple sample knowledge points. Compared with the prior art, the way of storing knowledge points and the way of retrieving knowledge points in the knowledge base are simple, and the question information can only be answered through a single knowledge point, resulting in low accuracy of information retrieval. When the information retrieval device needs to answer a question, the present application can answer it together through the sample knowledge points matched by the large model and the sample knowledge points associated with the matched sample knowledge points, thereby improving the accuracy of information retrieval. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0025] Figure 1 A schematic diagram of the structure of an information retrieval system provided in an embodiment of the present application;

[0026] Figure 2 A schematic diagram of the structure of an information retrieval device provided in an embodiment of the present application;

[0027] Figure 3 A flowchart of an information retrieval method provided in an embodiment of the present application;

[0028] Figure 4 A flowchart of another information retrieval method provided in an embodiment of the present application;

[0029] Figure 5 A flowchart of another information retrieval method provided in an embodiment of the present application;

[0030] Figure 6 A flowchart of another information retrieval method provided in an embodiment of the present application;

[0031] Figure 7 A flowchart of another information retrieval method provided in an embodiment of the present application;

[0032] Figure 8 A schematic diagram of a sample knowledge point relationship provided in an embodiment of the present application;

[0033] Fig. 9 A flowchart of another information retrieval method provided in an embodiment of the present application;

[0034] Fig.10 A schematic diagram of the structure of another information retrieval device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0036] The terms "first" and "second" and the like in the specification and drawings of this application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0037] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0038] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0039] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0040] In related technologies, large language models are based on deep learning technology, which requires a large number of samples and computing power, and a long period of training to achieve the ideal intelligent question-answering effect. When samples and computing power are insufficient, the Retrieval Augmented Generation (RAG) model is usually used to index and store knowledge. When a question is received, the question is answered with the help of the trained large language model.

[0041] Although the RAG model can answer questions from the knowledge base, due to the simple way of storing and retrieving knowledge points in the knowledge base, the RAG model cannot deeply understand the content of the knowledge points, resulting in low accuracy of information retrieval. Therefore, how to improve the accuracy of information retrieval is a technical problem that still needs to be solved.

[0042] In order to solve the technical problems existing in the related art, in an embodiment of the present application, the information retrieval device obtains multiple sample knowledge points, and obtains the semantics of each sample knowledge point in the multiple sample knowledge points based on the large language model. In this way, the information retrieval device can establish a relationship between any two sample knowledge points based on the semantics of each sample knowledge point in the multiple sample knowledge points. Compared with the prior art, the way of storing knowledge points and the way of retrieving knowledge points in the knowledge base are simple, and the question information can only be answered through a single knowledge point, resulting in low accuracy of information retrieval. When the information retrieval device needs to answer a question, the present application can answer it together through the sample knowledge points matched by the large model and the sample knowledge points associated with the matched sample knowledge points, thereby improving the accuracy of information retrieval.

[0043] Below, combined with the attached Figure 1 An information retrieval system 10 provided in an embodiment of the present application is described in detail. Figure 1 As shown, the information retrieval system 10 includes: an information retrieval device 11 and a terminal device 12.

[0044] Among them, the information retrieval device 11 is configured to: obtain multiple sample knowledge points; obtain the semantics of each of the sample knowledge points based on the multiple sample knowledge points and the large language model; and determine the relationship between any two sample knowledge points based on the semantics of each of the sample knowledge points.

[0045] The terminal device 12 is configured to send a plurality of sample knowledge points to the information retrieval apparatus 11 .

[0046] It should be noted that the terminal device 12 is a device with wireless communication function, which can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted. It can also be deployed on the water surface (such as ships, etc.). It can also be deployed in the air (for example, on airplanes, balloons and satellites, etc.). The user terminal is also called a mobile station (MS), a terminal (MT) and a terminal, etc., which is a device that provides voice and / or data connectivity to users. For example, the terminal includes a handheld device with wireless connection function, a vehicle-mounted device, etc. At present, the terminal can be: a mobile phone, a tablet computer, a laptop computer, a PDA, a mobile internet device (MID), a wearable device (such as a smart watch, a smart bracelet, a pedometer, etc.), a vehicle-mounted device (such as a car, a bicycle, an electric car, an airplane, a ship, a train, a high-speed rail, etc.), a virtual reality (VR) device, an augmented reality (AR) device, a wireless terminal in industrial control, a smart home device (such as a refrigerator, a TV, an air conditioner, an electric meter, etc.), an intelligent robot, a workshop device, a wireless terminal in self-driving, a wireless terminal in remote medical surgery, a wireless terminal in a smart grid, a wireless terminal in transportation safety, a wireless terminal in a smart city, or a wireless terminal in a smart home, a flying device (such as an intelligent robot, a hot air balloon, a drone, an airplane), etc. In a possible application scenario of the present application, the terminal device 12 is a terminal device that often works on the ground, such as a vehicle-mounted device. In this application, for the sake of convenience, chips deployed in the above-mentioned devices, such as system-on-a-chip (SOC), baseband chips, etc., or other chips with communication functions may also be referred to as user terminals.

[0047] Optionally, the terminal device 12 may be an embedded communication device or a user's handheld communication device, including a mobile phone, a tablet computer, etc.

[0048] As an example, in the embodiment of the present application, the user terminal 12 can also be a wearable device. Wearable devices can also be called wearable smart devices, which are a general term for wearable devices that use wearable technology to intelligently design and develop wearable devices for daily wear, such as glasses, gloves, watches, clothing and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also realize powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include full-featured, large-sized, and independent of smartphones to achieve complete or partial functions, such as smart watches or smart glasses, as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0049] The present application embodiment provides an information retrieval device for executing the information retrieval system provided by the present application embodiment. Figure 2 This is a schematic diagram of the structure of an information retrieval device provided in an embodiment of the present application. Figure 2 As shown, the information retrieval device 200 includes at least one processor 201, a communication line 202, and at least one communication interface 204, and may also include a memory 203. The processor 201, the memory 203 and the communication interface 204 may be connected via the communication line 202.

[0050] The processor 201 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application, such as one or more digital signal processors (DSP), or one or more field programmable gate arrays (FPGA).

[0051] The communication link 202 may include a pathway for transmitting information between the above-mentioned components.

[0052] The communication interface 204 is used to communicate with other devices or communication networks, and can use any transceiver-like device, such as Ethernet, radio access network (RAN), WLAN, etc.

[0053] The memory 203 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to include or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited to these.

[0054] In one possible design, the memory 203 can exist independently of the processor 201, that is, the memory 203 can be a memory outside the processor 201. In this case, the memory 203 can be connected to the processor 201 through the communication line 202, and is used to store execution instructions or application code, and the processor 201 controls the execution to implement the information retrieval method provided in the following embodiment of the present application. In another possible design, the memory 203 can also be integrated with the processor 201, that is, the memory 203 can be the internal memory of the processor 201, for example, the memory 203 is a high-speed cache, which can be used to temporarily store some data and instruction information, etc.

[0055] As an implementation method, the processor 201 may include one or more CPUs, for example Figure 2 As another possible implementation, the information retrieval device 200 may include multiple processors, such as Figure 2 As another implementation, the information retrieval apparatus 200 may further include an output device 205 and an input device 206.

[0056] Below, combined with the attached Figure 3 The information retrieval method provided in the embodiment of the present application is described in detail. Figure 3 As shown, the information retrieval method includes S301-S303.

[0057] S301. The information retrieval device obtains a plurality of sample knowledge points.

[0058] In a possible implementation, the information retrieval apparatus obtains a plurality of sample knowledge points from a terminal device.

[0059] Exemplarily, the sample knowledge point may be whole-house optical broadband terminal management requirements, and the sample knowledge point may also be whole-house optical broadband.

[0060] Optionally, the information retrieval device may be a server, and we do not limit this.

[0061] S302: The information retrieval device obtains the semantics of each sample knowledge point based on multiple sample knowledge points and a large language model.

[0062] In a possible implementation, the information retrieval device calls a large language model, and obtains the semantics of each sample knowledge point based on the large language model and multiple sample knowledge points.

[0063] In another possible implementation, the information retrieval device sends semantic recognition request information to the large language model, receives semantic recognition response information from the large language model, and determines the semantics of each sample knowledge point based on the semantic recognition response information.

[0064] Optionally, the semantic recognition request information includes multiple sample knowledge points, and the semantic recognition response information includes the semantics of each of the multiple sample knowledge points.

[0065] S303: The information retrieval device determines the relationship between any two sample knowledge points according to the semantics of each sample knowledge point.

[0066] In a possible implementation, the information retrieval device determines the semantic relationship and semantic similarity between any two sample knowledge points based on the semantics of each sample knowledge point.

[0067] The above scheme brings at least the following beneficial effects: In the embodiment of the present application, the information retrieval device obtains multiple sample knowledge points, and obtains the semantics of each sample knowledge point in the multiple sample knowledge points based on the large language model. In this way, the information retrieval device can establish a relationship between any two sample knowledge points based on the semantics of each sample knowledge point in the multiple sample knowledge points. Compared with the prior art, the way of storing knowledge points and the way of retrieving knowledge points in the knowledge base are simple, and the question information can only be answered through a single knowledge point, resulting in low accuracy of information retrieval. When the information retrieval device needs to answer a question, the present application can answer it together through the sample knowledge points matched by the large model and the sample knowledge points associated with the matched sample knowledge points, thereby improving the accuracy of information retrieval.

[0068] In one possible implementation, combining Figure 3 ,like Figure 4As shown, in S303, after the information retrieval device determines the relationship between any two sample knowledge points according to the semantics of each sample knowledge point, the information retrieval device stores the sample knowledge points. The process of the information retrieval device storing the sample knowledge points can be specifically implemented by the following S401.

[0069] S401. The information retrieval device stores the sample knowledge point and the relationship between any two sample knowledge points into a vector library.

[0070] In a possible implementation, the information retrieval device vectorizes the sample knowledge point and the relationship between any two sample knowledge points, and stores the vectorized sample knowledge point and the relationship between any two sample knowledge points in a vector library.

[0071] It should be explained that the information retrieval device can call the vector library to answer the questions of the terminal device.

[0072] The above scheme brings at least the following beneficial effects: In the embodiment of the present application, the information retrieval device vectorizes the sample knowledge point and the relationship between any two sample knowledge points, and stores the vectorized sample knowledge point and the relationship between any two sample knowledge points in the vector library. In this way, when the information retrieval device receives a question request from a terminal device, it can call the vector library to achieve the purpose of quickly answering the question request.

[0073] In one possible implementation, combining Figure 3 ,like Figure 5 As shown, in S303, after the information retrieval device determines the relationship between any two sample knowledge points according to the semantics of each sample knowledge point, the information retrieval device generates retrieval response information. The process of the information retrieval device generating retrieval response information can be specifically implemented through the following S501-S505.

[0074] S501: The information retrieval apparatus receives retrieval request information from a terminal device.

[0075] The search request information includes at least one keyword.

[0076] In a possible implementation, after receiving the search request information from the terminal device, the information search apparatus performs vectorization processing on at least one keyword in the search request information.

[0077] It should be explained that the terminal device collects surrounding sounds through a sound collection device, performs noise reduction processing on the collected sounds, generates text information based on the sound information, and sends the text information to the information retrieval device.

[0078] Optionally, the text information includes at least one keyword.

[0079] S502: The information retrieval device determines a first sample knowledge point corresponding to at least one keyword based on at least one keyword and a large language model.

[0080] Among them, the first sample knowledge point is a sample knowledge point among multiple sample knowledge points.

[0081] In a possible implementation, the information retrieval device calls a large language model to determine a first sample knowledge point that is close to an answer to at least one keyword from a plurality of sample knowledge points.

[0082] S503: The information retrieval device obtains a first relationship from the relationship between any two sample knowledge points.

[0083] The first relationship is a relationship between the first sample knowledge point and sample knowledge points other than the first sample knowledge point among the plurality of sample knowledge points.

[0084] Optionally, the first relationship includes a first semantic relationship and a first semantic similarity. The first semantic relationship includes a semantic relationship between the first sample knowledge point and a sample knowledge point other than the first sample knowledge point among the multiple sample knowledge points, and the first semantic similarity includes a semantic similarity between the first sample knowledge point and a sample knowledge point other than the first sample knowledge point among the multiple sample knowledge points.

[0085] S504: The information retrieval device determines a second sample knowledge point based on at least one keyword, the first sample knowledge point and the first relationship.

[0086] In a possible implementation, the information retrieval device determines a second sample knowledge point having the greatest correlation with the at least one keyword based on the at least one keyword, the first sample knowledge point and the first semantic relationship.

[0087] In another possible implementation, the information retrieval device determines a second sample knowledge point having the highest semantic similarity to the first sample knowledge point based on at least one keyword, the first sample knowledge point and the first semantic relationship.

[0088] S505: The information retrieval device generates retrieval response information based on the first sample knowledge point and the second sample knowledge point.

[0089] In a possible implementation, the information retrieval device generates retrieval response information based on the first sample knowledge point, the second sample knowledge point and the large language model.

[0090] The above scheme brings at least the following beneficial effects: in an embodiment of the present application, when the information retrieval device needs to answer a question, the present application can answer it jointly through the first sample knowledge point matched by the large model and the second sample knowledge point associated with the matched sample knowledge point, thereby improving the accuracy of information retrieval.

[0091] In one possible implementation, combining Figure 5 ,like Figure 6 As shown, S504, the process in which the information retrieval device determines the second sample knowledge point based on at least one keyword, the first sample knowledge point and the first relationship can be specifically implemented through the following S601-S602.

[0092] S601: The information retrieval device generates a key sentence based on at least one keyword.

[0093] In a possible implementation, the information retrieval device generates a key sentence based on at least one keyword, context, and context.

[0094] S602: The information retrieval device determines a second sample knowledge point based on the key sentence, the first sample knowledge point and the first relationship.

[0095] In a possible implementation, the information retrieval device determines, from a plurality of knowledge points, a second sample knowledge point having the greatest correlation with at least one keyword based on the key sentence, the first sample knowledge point and the first semantic relationship.

[0096] In another possible implementation, the information retrieval device determines a second sample knowledge point having the highest semantic similarity to the first sample knowledge point based on the key sentence, the first sample knowledge point and the first semantic relationship.

[0097] The above scheme brings at least the following beneficial effects: In the embodiment of the present application, the information retrieval device generates a key sentence based on at least one keyword, context and context, so that the information retrieval device can better determine the second sample knowledge point.

[0098] In one possible implementation, combining Figure 5 ,like Figure 7 As shown, S505, the process in which the information retrieval device generates retrieval response information based on the first sample knowledge point and the second sample knowledge point can be specifically implemented through the following S701-S704.

[0099] S701. The information retrieval device determines whether the first sample knowledge point and the second sample knowledge point satisfy a condition for generating retrieval response information.

[0100] Optionally, a condition for generating the retrieval response information is that the total amount of the first sample knowledge point and the second sample knowledge point is greater than a first threshold.

[0101] S702: When the first sample knowledge point and the second sample knowledge point do not satisfy the condition for generating search response information, the information search apparatus obtains a second relationship of the second sample knowledge point.

[0102] The second relationship is the relationship between the second sample knowledge point and sample knowledge points other than the second sample knowledge point among the multiple sample knowledge points.

[0103] Optionally, the second relationship includes a second semantic relationship and a second semantic similarity. The second semantic relationship includes a semantic relationship between the second sample knowledge point and sample knowledge points other than the second sample knowledge point among the multiple sample knowledge points, and the second semantic similarity includes a semantic similarity between the second sample knowledge point and sample knowledge points other than the second sample knowledge point among the multiple sample knowledge points.

[0104] S703: The information retrieval device determines a third sample knowledge point based on at least one keyword, the second sample knowledge point and the second relationship.

[0105] In a possible implementation, the information retrieval device determines a third sample knowledge point having the greatest correlation with the at least one keyword based on the at least one keyword, the second sample knowledge point and the second semantic relationship.

[0106] In another possible implementation, the information retrieval device determines a third sample knowledge point having the highest semantic similarity to the second sample knowledge point based on at least one keyword, the second sample knowledge point and the second semantic relationship.

[0107] For example, Figure 8 As shown, the first sample knowledge point is: Whole-house optical broadband, and the sample knowledge points associated with the first sample knowledge point are: FTTR, whole-house optical broadband business acceptance, whole-house optical broadband terminal management requirements, and whole-house optical broadband business specifications. The relationship between the first sample knowledge point and FTTR is that the English abbreviation of whole-house optical broadband is FTTR; the relationship between the first sample knowledge point and the whole-house optical broadband terminal management requirements is to describe the whole-house optical broadband business acceptance process; the relationship between the first sample knowledge point and the whole-house optical broadband business acceptance is to describe the whole-house optical broadband terminal management requirements; the relationship between the first sample knowledge point and the whole-house optical broadband business specifications is that the business management specifications mainly describe the relevant provisions of the whole-house optical bandwidth. The second sample knowledge point is: Whole-house optical broadband business specifications, and the third sample knowledge point is: 08 Value-added Business Management Specifications (49 books), and the second relationship is that the whole-house optical bandwidth management specifications belong to the value-added business management specifications.

[0108] S704: The information retrieval device generates retrieval response information based on the first sample knowledge point, the second sample knowledge point and the third sample knowledge point.

[0109] In a possible implementation, the information retrieval device generates retrieval response information based on the first sample knowledge point, the second sample knowledge point, the third sample knowledge point and the large language model.

[0110] The above scheme brings at least the following beneficial effects: In the embodiment of the present application, when the first sample knowledge point and the second sample knowledge point do not meet the conditions for generating retrieval response information, the information retrieval device obtains the second relationship of the second sample knowledge point, so that the information retrieval device can determine the third sample knowledge point based on at least one keyword, the second sample knowledge point and the second relationship. The retrieval response information is generated by the third sample knowledge point associated with the second sample knowledge point, thereby improving the accuracy of replying to the question.

[0111] In one possible implementation, combining Figure 3 ,like Fig. 9 As shown, S302, the process in which the information retrieval device obtains the semantics of each sample knowledge point based on multiple sample knowledge points and a large language model can be specifically implemented through the following S901-S902.

[0112] S901. The information retrieval device inputs a plurality of sample knowledge points into a large language model to obtain a prediction result.

[0113] Among them, the prediction results are used to characterize the semantics of each sample knowledge point.

[0114] S902: When the content volume of the prediction result is less than a threshold, the information retrieval device responds to the sample marking operation on each sample knowledge point and determines the semantics corresponding to the marking operation as the semantics of the sample knowledge point.

[0115] It should be explained that when the content of the prediction result is less than the threshold, the large language model cannot accurately identify the semantics of each sample knowledge point. In this way, the information retrieval device can be displayed to the manual service platform, which performs sample marking operations on each sample knowledge point and determines the semantics corresponding to the marking operation provided by the manual service platform as the semantics of the sample knowledge point, thereby improving the recognition efficiency of the semantics of the sample knowledge point.

[0116] The above scheme brings at least the following beneficial effects: In the embodiment of the present application, when the large language model cannot accurately identify multiple sample knowledge points, the artificial service platform can accurately add semantics to the multiple sample knowledge points to facilitate the establishment of the relationship between the sample knowledge points.

[0117] It can be seen that the technical solution provided by the embodiment of the present application is introduced above mainly from the perspective of the method. In order to realize the above functions, it includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should be easily aware that, in combination with the modules and algorithm steps of each example described in the embodiment disclosed herein, the embodiment of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0118] An embodiment of the present application also provides a computer-readable storage medium, in which instructions are stored. When a computer executes the instructions, the computer executes each step in the method flow shown in the above method embodiment.

[0119] like Fig.10 The figure is a schematic diagram of the structure of an information retrieval device provided in an embodiment of the present application. The information retrieval device can be used to perform Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 , Fig. 9 The information retrieval method shown. Fig.10 The information retrieval device 100 shown includes: an acquisition unit 1001 , a processing unit 1002 , and a communication unit 1003 .

[0120] The acquisition unit 1001 is used to acquire multiple sample knowledge points; the acquisition unit 1001 is also used to acquire the semantics of each of the sample knowledge points based on the multiple sample knowledge points and the large language model; the processing unit 1002 is used to determine the relationship between any two sample knowledge points based on the semantics of each sample knowledge point; the relationship between any two sample knowledge points is used to retrieve the sample knowledge points.

[0121] Optionally, the processing unit 1002 is further configured to store the sample knowledge point and the relationship between any two sample knowledge points in a vector library.

[0122] Optionally, the communication unit 1003 is used to receive retrieval request information from a terminal device; the retrieval request information includes at least one keyword; the processing unit 1002 is also used to determine a first sample knowledge point corresponding to the at least one keyword based on the at least one keyword and the large language model; the first sample knowledge point is a sample knowledge point among the multiple sample knowledge points; the processing unit 1002 is also used to obtain a first relationship from the relationship between any two sample knowledge points; the first relationship is a relationship between the first sample knowledge point and a knowledge point among the multiple sample knowledge points other than the first sample knowledge point; the processing unit 1002 is also used to determine a second sample knowledge point based on the at least one keyword, the first sample knowledge point and the first relationship; the processing unit 1002 is also used to generate retrieval response information based on the first sample knowledge point and the second sample knowledge point.

[0123] Optionally, the processing unit 1002 is further used to generate a key sentence based on the at least one keyword; the processing unit 1002 is further used to determine the second sample knowledge point based on the key sentence, the first sample knowledge point and the first relationship.

[0124] Optionally, the processing unit 1002 is further used to determine whether the first sample knowledge point and the second sample knowledge point satisfy the condition for generating the retrieval response information; the processing unit 1002 is further used to instruct the acquisition unit 1001 to acquire the second relationship of the second sample knowledge point when the first sample knowledge point and the second sample knowledge point do not satisfy the condition for generating the retrieval response information; the second relationship is the relationship between the second sample knowledge point and the knowledge points among the multiple sample knowledge points except the second sample knowledge point; the processing unit 1002 is further used to determine the third sample knowledge point based on the at least one keyword, the second sample knowledge point and the second relationship; the processing unit 1002 is further used to generate the retrieval response information based on the first sample knowledge point, the second sample knowledge point and the third sample knowledge point.

[0125] Optionally, the processing unit 1002 is further used to input the multiple sample knowledge points into the large language model to obtain a prediction result; the prediction result is used to characterize the semantics of each sample knowledge point; the processing unit 1002 is also used to, in response to a sample marking operation on each of the sample knowledge points, determine the semantics corresponding to the marking operation as the semantics of the sample knowledge point when the content volume of the prediction result is less than a threshold.

[0126] An embodiment of the present application also provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instruction to implement the information retrieval method in the above method embodiment.

[0127] An embodiment of the present application provides a computer program product including instructions. When the instructions are executed on a computer, the computer is enabled to execute the industrial terminal switching method in the above method embodiment.

[0128] Among them, the computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, and a hard disk. Random Access Memory (RAM), Read-Only Memory (ROM), Erasable Programmable Read Only Memory (EPROM), registers, hard disks, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any other form of computer-readable storage medium in a suitable combination of the above, or numerical values ​​in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an Application Specific Integrated Circuit (ASIC). In embodiments of the present invention, computer-readable storage media may be any tangible media that contains or stores a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0129] Since the apparatus, device, computer-readable storage medium, and computer program product in the embodiments of the present invention can be applied to the above-mentioned method, the technical effects that can be obtained can also refer to the above-mentioned method embodiments, and the embodiments of the present application will not be repeated here.

[0130] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An information retrieval method, characterized in that: The method comprises: Obtain multiple sample knowledge points; According to the multiple sample knowledge points and the large language model, the semantics of each of the sample knowledge points are obtained; according to the semantics of each of the sample knowledge points, the relationship between any two sample knowledge points is determined; the relationship between any two sample knowledge points is used to retrieve the sample knowledge points.

2. The method according to claim 1, characterized in that After determining the relationship between any two sample knowledge points according to the semantics of each of the sample knowledge points, the method further includes: The sample knowledge point and the relationship between any two sample knowledge points are stored in a vector library.

3. The method according to claim 2, characterized in that After determining the relationship between any two sample knowledge points according to the semantics of each of the sample knowledge points, the method further includes: Receiving a search request message from a terminal device; the search request message includes at least one keyword; Based on the at least one keyword and the large language model, determining a first sample knowledge point corresponding to the at least one keyword; the first sample knowledge point is a sample knowledge point among the multiple sample knowledge points; Acquire a first relationship from the relationship between any two sample knowledge points; the first relationship is a relationship between the first sample knowledge point and sample knowledge points other than the first sample knowledge point among the plurality of sample knowledge points; Determining a second sample knowledge point based on the at least one keyword, the first sample knowledge point, and the first relationship; Generate retrieval response information based on the first sample knowledge point and the second sample knowledge point.

4. The method according to claim 3, characterized in that: The determining the second sample knowledge point based on the at least one keyword, the first sample knowledge point and the relationship between any two sample knowledge points includes: generating a key sentence based on the at least one keyword; Based on the key sentence, the first sample knowledge point and the first relationship, the second sample knowledge point is determined.

5. The method according to claim 3, characterized in that: The generating retrieval response information based on the first sample knowledge point and the second sample knowledge point includes: Determining whether the first sample knowledge point and the second sample knowledge point meet the condition for generating the search response information; When the first sample knowledge point and the second sample knowledge point do not satisfy the condition for generating the retrieval response information, obtaining a second relationship of the second sample knowledge point; the second relationship is a relationship between the second sample knowledge point and sample knowledge points other than the second sample knowledge point among the plurality of sample knowledge points; Determining a third sample knowledge point based on the at least one keyword, the second sample knowledge point, and the second relationship; The search response information is generated based on the first sample knowledge point, the second sample knowledge point and the third sample knowledge point.

6. The method according to any one of claims 1 to 5, characterized in that: The acquiring the semantics of each of the sample knowledge points according to the plurality of sample knowledge points and the large language model comprises: Inputting the plurality of sample knowledge points into the large language model to obtain a prediction result; the prediction result is used to characterize the semantics of each of the sample knowledge points; In a case where the content volume of the prediction result is less than a threshold, in response to a sample marking operation on each of the sample knowledge points, the semantics corresponding to the marking operation is determined as the semantics of the sample knowledge point.

7. An information retrieval device, characterized in that: The device comprises: an acquisition unit and a processing unit; An acquisition unit, used for acquiring multiple sample knowledge points; The acquisition unit is used to acquire the semantics of each of the sample knowledge points according to the multiple sample knowledge points and the large language model; The processing unit is used to determine the relationship between any two sample knowledge points according to the semantics of each of the sample knowledge points; the relationship between any two sample knowledge points is used to retrieve the sample knowledge points.

8. An information retrieval device, characterized in that: include: A processor and a memory; wherein the memory is used to store computer-executable instructions, and when the information retrieval device is running, the processor executes the computer-executable instructions stored in the memory, so that the information retrieval device executes the information retrieval method described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium comprises instructions, which, when executed by an information retrieval device, cause the computer to execute the information retrieval method according to any one of claims 1 to 6.

10. A computer program product, comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the information retrieval method described in any one of claims 1 to 6 is implemented.

Images