Practical learning information query method and device, electronic equipment and medium

Abstract

Embodiments of the present disclosure disclose a practice learning information query method, device, electronic equipment and medium. A specific embodiment of the method includes: obtaining practice learning information query request information; inputting the practice learning query request information into a practice learning information query large model to obtain a query result information sequence; in response to determining that the query result information sequence meets a preset sending condition, optimizing each query result information sequence in the query result information sequence to generate query summary information, obtaining a query summary information sequence; and sending the query result information sequence to a first user terminal. This embodiment can reduce the waste of time for the user terminal to query practice learning information and improve user experience.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, and more specifically to practical learning information retrieval methods, apparatus, electronic devices, and media. Background Technology

[0002] During the headhunting practice learning phase, when questions arise, it is often necessary to sift through and search through a large amount of learning materials. Currently, the common methods for searching relevant learning materials are: directly searching for relevant names in structured databases or file systems, or using a large language model to sequentially search for keywords in each document in the knowledge base, directly selecting documents containing the keywords of the query information as the retrieved learning materials.

[0003] However, in practice, it has been found that when using the above method to retrieve practical learning information, the following technical problems often arise:

[0004] First, in a structured database or file system, searching for practical learning information solely by the name of the learning material is insufficient because the name of the learning material cannot accurately summarize its content. This leads to a decrease in the accuracy of the retrieved practical learning information, requiring users to repeatedly search for it, thus wasting their time and resulting in a poor user experience.

[0005] Second, the method of sequentially searching for keywords in each document in the knowledge base and then outputting the document containing the keywords of the query information will increase the time required for retrieval as the size of the knowledge base expands, making it difficult to retrieve practical learning information in a timely manner.

[0006] In the process of adopting technical solutions to address the aforementioned technical problems, the following issues often arise: When the large language model fails to retrieve practical learning information from the knowledge base, it may randomly generate text as practical learning information, leading to a decrease in the accuracy of the output practical learning information. A conventional solution to this problem typically involves setting instructions to prevent the large language model from randomly generating text, sending query requests to a query server to alert associated second-user terminals, and receiving practical learning information uploaded by user terminals from each second-user terminal. However, this conventional solution still suffers from the following third technical problem: When sending queries to each associated second-user terminal via the query server, the first-user terminal must wait for a response from each second-user terminal. The practical learning information responded by some second-user terminals may have low accuracy, resulting in a decrease in the accuracy of the obtained practical learning information. This wastes the time the first-user terminal spends querying practical learning information, leading to a poor user experience.

[0007] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0008] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0009] Some embodiments of this disclosure provide practical learning information retrieval methods, apparatuses, electronic devices, and media to address one or more of the technical problems mentioned in the background section above.

[0010] In a first aspect, some embodiments of this disclosure provide a method for querying practical learning information. The method includes: acquiring practical learning information query request information; inputting the practical learning query request information into a large-scale practical learning information query model to obtain a query result information sequence; in response to determining that the query result information sequence meets preset sending conditions, optimizing each query result information sequence in the query result information sequence to generate query summary information, and obtaining a query summary information sequence; and sending the query result information sequence to a first user terminal.

[0011] Secondly, some embodiments of this disclosure provide a practical learning information query device, the device comprising: an acquisition unit configured to acquire practical learning information query request information; an input unit configured to input the practical learning query request information into a practical learning information query big model to obtain a query result information sequence; an optimization unit configured to, in response to determining that the query result information sequence meets preset sending conditions, optimize each query result information sequence in the query result information sequence to generate query summary information, thereby obtaining a query summary information sequence; and a sending unit configured to send the query result information sequence to a first user terminal.

[0012] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any of the implementations of the first aspect above.

[0013] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0014] The above embodiments of this disclosure have the following beneficial effects: the practical learning information query method of some embodiments of this disclosure can reduce the wasted time of user terminals querying practical learning information and improve the user experience. Specifically, the reason for wasting user terminals' time querying practical learning information and resulting in a poor user experience is that in a structured database or file system, querying practical learning information only by searching for the name of the learning material is difficult because the name of the learning material cannot accurately summarize the content of the learning material, resulting in a decrease in the accuracy of the retrieved practical learning information, and the user terminal needs to repeatedly query practical learning information. Based on this, the practical learning information query method of some embodiments of this disclosure first obtains practical learning information query request information. Second, the above practical learning query request information is input into a pre-trained practical learning information query big model to obtain a query result information sequence. Thus, practical learning information can be initially retrieved through the big language model. Then, in response to determining that the above query result information sequence meets the preset sending conditions, each query result information sequence in the above query result information sequence is optimized to generate query summary information, resulting in a query summary information sequence. Thus, practical learning information that is easier for user terminals to understand can be obtained. Finally, the above query result information sequence is sent to the first user terminal. Therefore, the first user terminal can receive the queried practical learning information. Thus, some practical learning information query methods disclosed herein can query practical learning information from a knowledge base stored in the cloud or locally using a large model. Then, the queried information can be optimized to send the optimized practical learning information to the user terminal. This allows for the retrieval of learning materials from the knowledge base and the use of related learning materials as queried practical learning information. This improves the accuracy of the retrieved practical learning information, thereby reducing the number of times the user terminal repeatedly queries practical learning information, reducing wasted time on the user terminal, and improving the user experience. Attached Figure Description

[0015] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0016] Figure 1 This is a flowchart of some embodiments of the practical learning information query method according to this disclosure;

[0017] Figure 2 These are schematic diagrams illustrating the structure of some embodiments of the practical learning information query device according to this disclosure;

[0018] Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0019] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0020] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0021] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0022] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0023] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0024] Before performing any of the operations involving the collection, storage, or use of user personal information disclosed in this disclosure, the relevant organizations or individuals shall fulfill their obligations, including conducting personal information security impact assessments, informing personal information subjects, and obtaining prior authorization and consent from personal information subjects.

[0025] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] Figure 1 A flow 100 of some embodiments of the practical learning information retrieval method according to the present disclosure is shown. The practical learning information retrieval method includes the following steps:

[0027] Step 101: Obtain the practical learning information query request information.

[0028] In some embodiments, the entity executing the practical learning information query method can obtain the practical learning information query request information from a first user terminal via a wired or wireless connection. The first user terminal can be the terminal that wants to query the practical learning information.

[0029] As an example, the above-mentioned practical learning information query request information may be, but is not limited to, at least one of the following: information representing "What skills are needed to learn in the headhunting industry?", information representing "What are the techniques for reading resumes?", or information representing "How to screen resumes?".

[0030] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other currently known or future wireless connection methods.

[0031] Step 102: Input the practical learning query request information into the practical learning information query big model to obtain the query result information sequence.

[0032] In some embodiments, the aforementioned executing entity may input the aforementioned practical learning query request information into a large-scale practical learning information query model to obtain a sequence of query result information. The aforementioned large-scale practical learning information query model may be a neural network model that takes the practical learning query request information as input and the sequence of query result information as output.

[0033] As an example, the above-mentioned practical learning information query model can be an LLM (Large Language Model).

[0034] In some optional implementations of certain embodiments, the execution entity inputs the aforementioned practical learning query request information into the practical learning information query big model to obtain a query result information sequence, which may include the following steps:

[0035] The first step is to obtain the practical learning knowledge base set. This set can be obtained from a storage terminal or the cloud. The storage terminal can be a terminal used to store the practical learning knowledge base set. The practical learning knowledge base within the set can be a knowledge base used to store relevant knowledge about practical learning. This practical learning knowledge base may include: a set of practical learning files.

[0036] As an example, the relevant knowledge learned in the above practical learning can be relevant knowledge learned in the target industry. The target industry could be the headhunting industry. The format of the practical learning files in the above practical learning document set can be, but is not limited to, at least one of the following: documents, presentations, or videos.

[0037] The second step involves transforming each practical learning knowledge base in the aforementioned practical learning knowledge base set to generate a practical learning knowledge graph, resulting in a practical learning knowledge graph set. The practical learning knowledge graphs in this set may include, but are not limited to, a practical learning knowledge vector set. Here, a preset transformation algorithm can be used to transform each practical learning knowledge base in the aforementioned practical learning knowledge base set to generate a practical learning knowledge graph, resulting in a practical learning knowledge graph set. Each practical learning knowledge graph in the aforementioned practical learning knowledge vector set can be a vector database. The practical learning knowledge vectors in the aforementioned practical learning knowledge vector set can represent a character or a word from the aforementioned practical learning knowledge base.

[0038] As an example, the aforementioned preset conversion algorithm may be, but is not limited to, at least one of the following: ERNIE-Embedding (Enhanced Language Representation with Informative Entities) algorithm, M3E (Moka Massive Mixed Embedding) algorithm, and BGE (BAAI General Embedding) algorithm.

[0039] Third, based on the aforementioned practical learning query request information, a first matching process is performed on the aforementioned practical learning knowledge base set to obtain a first matching knowledge information set. This first matching process can be performed using a preset first matching algorithm. The first matching knowledge information in the aforementioned first matching knowledge information set may be a fragment of a practical learning file from the practical learning file set included in the practical learning knowledge base set.

[0040] As an example, the first matching algorithm mentioned above could be the BM25 (optimal matching) algorithm.

[0041] Fourth, based on the above-mentioned practice learning knowledge graph set and the above-mentioned practice learning query request information, the above-mentioned first matching knowledge information set is filtered to obtain the filtered knowledge information set.

[0042] Fifth, based on the above-mentioned practical learning query request information, perform a second matching process on the filtered knowledge information set to obtain a second matched knowledge information set. This second matching process can be performed using a preset second matching algorithm.

[0043] As an example, the second matching algorithm mentioned above could be a cross-encoder model algorithm.

[0044] Step 6: Merge the aforementioned practical learning query request information and the aforementioned second matching knowledge information set to obtain model query instruction information. Specifically, the fusion of the aforementioned practical learning query request information and the aforementioned second matching knowledge information set to obtain model query instruction information can be achieved by determining the aforementioned practical learning query request information and the aforementioned second matching knowledge information set as the practical learning query request information and the second matching knowledge information set included in the aforementioned model query instruction information.

[0045] As an example, the query instruction information for the above model could be a prompt.

[0046] Step 7: Input the above model query command information into the above practical learning information query big model to obtain the above query result information sequence.

[0047] In some optional implementations of certain embodiments, the execution entity filters the first matching knowledge information set based on the practice learning knowledge graph set and the practice learning query request information to obtain a filtered knowledge information set, which may include the following steps:

[0048] The first step is to vectorize the above-mentioned practical learning query request information to obtain a query request vector. This can be achieved using the aforementioned preset conversion algorithm.

[0049] The second step is to perform the following transformation sub-steps for each piece of first matching knowledge information in the aforementioned first matching knowledge information set:

[0050] The first sub-step involves determining the practice learning knowledge graph corresponding to the first matching knowledge information in the aforementioned practice learning knowledge graph set as the target knowledge graph. This target knowledge graph may include, but is not limited to, a practice learning knowledge vector set. Determining the practice learning knowledge graph corresponding to the first matching knowledge information in the aforementioned practice learning knowledge graph set as the target knowledge graph can be achieved by determining the practice learning knowledge graph corresponding to the practice learning knowledge base corresponding to the first matching knowledge information in the aforementioned practice learning knowledge graph set as the target knowledge graph.

[0051] The second sub-step involves determining the target knowledge vector set as the set of practical learning knowledge vectors that correspond to the first matching knowledge information within the aforementioned target knowledge graph. Specifically, the practical learning knowledge vectors corresponding to each character in the fragment of the practical learning file, as described above, can be determined as target knowledge vectors, thus obtaining the aforementioned target knowledge vector set.

[0052] The third sub-step involves concatenating the target knowledge vectors in the aforementioned target knowledge vector set to obtain a concatenated target knowledge vector. Specifically, the concatenation can be performed according to the order of the characters in the first matching knowledge information to obtain the concatenated target knowledge vector.

[0053] The fourth sub-step involves determining the query similarity value between the query request vector and the target knowledge concatenation vector. This can be achieved using a pre-defined similarity algorithm.

[0054] As an example, the aforementioned preset similarity algorithm may be, but is not limited to, at least one of the following: cosine similarity algorithm, Euclidean distance algorithm, or Manhattan distance algorithm.

[0055] The third step is to sort the determined query similarity values ​​to obtain a sequence of query similarity values. This can be done by sorting the determined query similarity values ​​in descending order.

[0056] The fourth step is to determine the target number of query similarity values ​​and the corresponding first matching knowledge information of the above query similarity value sequence as the filtering knowledge information, thus obtaining the above filtering knowledge information set.

[0057] As an example, the target quantity mentioned above can be, but is not limited to, at least one of the following: 10, 20, 30.

[0058] The content related to step 102, as an inventive point of this disclosure, solves the second technical problem mentioned in the background art: "difficulty in timely retrieval of practical learning information." The factors leading to this difficulty often include: sequentially searching for keywords in each document of the knowledge base and then outputting the document containing the keywords of the query information; as the size of the knowledge base increases, the retrieval time increases. Solving these factors would allow for timely retrieval of practical learning information. To achieve this, this disclosure can first perform a preliminary search of the knowledge base to obtain keywords with lower accuracy. Then, based on the vectorized knowledge base, the keywords obtained from the preliminary search can be filtered to improve their accuracy. Next, the filtered keywords can be further matched and searched to obtain keywords with higher accuracy from the knowledge base. Finally, the obtained keywords and query information can be input into a large model for further retrieval of the knowledge base. This improves the accuracy of the retrieved practical learning information and ensures timely retrieval.

[0059] Step 103: In response to determining that the query result information sequence meets the preset sending conditions, each query result information in the query result information sequence is optimized to generate query summary information, thus obtaining the query summary information sequence.

[0060] In some embodiments, the execution entity may, in response to determining that the query result information sequence meets a preset sending condition, optimize each query result information in the query result information sequence to generate query summary information, thereby obtaining a query summary information sequence. This optimization can be performed using a preset optimization algorithm. The preset sending condition may be that the query result information sequence is not empty.

[0061] As an example, the aforementioned preset optimization algorithm could be an LLM (Large Language Model) algorithm.

[0062] Optionally, the aforementioned implementing entity may also perform the following steps:

[0063] The first step involves generating a practical learning reminder message based on the practical learning information query request message, in response to the determination that the query result information sequence does not meet the preset sending conditions. This practical learning reminder message can be used to remind the second user terminal to receive the practical learning information query request message. The second user terminal can be a terminal used to receive the practical learning reminder message.

[0064] The second step involves, in response to receiving the reminder message sending instruction from the first user terminal, sending the practical learning reminder message to the practical learning server and each of the corresponding second user terminals. The practical learning server may be a server used for communication between the first user terminal and each of the second user terminals.

[0065] The third step involves receiving practical learning information from each of the aforementioned second user terminals to obtain a practical learning information set, and then sending the practical learning information set to the aforementioned first user terminal. This can be achieved through the aforementioned practical learning server, which receives practical learning information from each of the aforementioned second user terminals to obtain the practical learning information set and then sends the practical learning information set to the aforementioned first user terminal. The aforementioned practical learning information may be practical learning materials generated by the aforementioned second user terminals.

[0066] Considering the problems with the conventional solutions mentioned above, and addressing the third technical issue—that the accuracy of practical learning information responses from some user terminals is low when the query server sends data to each associated user terminal, thus reducing the overall accuracy of the obtained practical learning information—the following solution can be adopted based on the available technology.

[0067] In some optional implementations of certain embodiments, the execution entity generates practical learning reminder information based on the practical learning information query request information, which may include the following steps:

[0068] The first step is to obtain the second user basic information set. This second user basic information set may include, but is not limited to, at least one of the following: a second user terminal identifier and a second user terminal profile document. The second user terminal identifier can uniquely identify a second user terminal.

[0069] As an example, the aforementioned second user terminal introduction document could be a "personal instruction manual".

[0070] The second step involves inputting each piece of second user basic information from the aforementioned second user basic information set into a pre-trained user feature information extraction model to generate second user feature information, thus obtaining the second user feature information set. Specifically, the second user terminal introduction document included in the aforementioned second user basic information can be input into the pre-trained user feature information extraction model to generate the second user feature information. The aforementioned pre-trained user feature information extraction model can be a pre-trained neural network model that takes the second user terminal introduction document as input and the second user feature information as output.

[0071] The third step involves matching the aforementioned practical learning information query request information with each second user feature information in the aforementioned second user feature information set to generate second user matching information, thus obtaining a second user matching information set. Specifically, the aforementioned practical learning query request information can be vectorized using the aforementioned preset conversion algorithm to obtain a query request vector. Then, for each second user feature information in the aforementioned second user feature information set, the following generation sub-steps can be performed: First, the aforementioned preset similarity algorithm can be used to determine the second user similarity value corresponding to the query request vector and the aforementioned second user feature information. Second, in response to determining that the second user similarity value is greater than or equal to the target similarity value, the information representing "successful matching" is determined as the aforementioned second user matching information. Third, in response to determining that the second user similarity value is less than the target similarity value, the information representing "unsuccessful matching" is determined as the aforementioned second user matching information. Finally, the generated second user matching information can be determined as the aforementioned second user matching information set.

[0072] As an example, the target similarity value mentioned above could be 0.8.

[0073] The fourth step is to determine the basic information of the second user corresponding to the second user matching information in the above-mentioned second user matching information set that meets the preset matching conditions as the target second user basic information, thus obtaining the target second user basic information set. Here, the preset matching conditions can be that the second user matching information is information representing "successful matching".

[0074] Fifth, the aforementioned target second user basic information set and the aforementioned practical learning information query request information are fused together to obtain the aforementioned practical learning reminder information. Specifically, the fusion of the aforementioned target second user basic information set and the aforementioned practical learning information query request information to obtain the aforementioned practical learning reminder information can be achieved by defining the aforementioned target second user basic information set and the aforementioned practical learning information query request information as the target second user basic information set and practical learning information query request information included in the aforementioned practical learning reminder information.

[0075] Optionally, before inputting each piece of second user basic information from the aforementioned second user basic information set into the pre-trained user feature information extraction model to generate second user feature information, the aforementioned execution entity may also perform the following steps:

[0076] The first step is to obtain the basic information set of sample users. This basic information set can be obtained from the aforementioned storage terminal. The basic information in this set may include, but is not limited to, at least one of the following: a user terminal identifier and a user terminal description document. The user terminal identifier uniquely identifies the basic information of a user terminal. The user terminal description document represents the characteristic information of a user terminal.

[0077] The second step involves selecting basic information of sample users from the aforementioned set of basic information and performing the following training sub-steps:

[0078] The first sub-step involves inputting the basic information of the sample users into the vector transformation sub-model included in the initial user feature information extraction model to obtain the initial user base vector. This initial user feature information extraction model further includes a feature extraction sub-model and a feature annotation sub-model. The vector transformation sub-model maps each input character to a word vector, the feature extraction sub-model extracts the semantic information of the context between the word vectors, and the feature annotation sub-model further extracts features from the extracted semantic information. Sample user basic information can be randomly selected from the aforementioned set of sample user basic information. The initial user feature information extraction model can be an untrained neural network model that takes a user terminal profile document (including the sample user basic information) as input and outputs user feature information.

[0079] Specifically, the aforementioned vector transformation sub-model can be divided into two layers: a character embedding layer and an encoding layer. The vector transformation sub-model, which inputs the basic information of the sample user into the initial user feature information extraction model to obtain the initial user base vector, can be as follows: For each character in the user terminal profile document included in the basic information of the sample user, the character is input into the aforementioned character embedding layer to obtain a character vector, a text vector, and a position vector. Then, the sum of the aforementioned character vector, text vector, and position vector can be determined as the initial character feature vector. Next, the obtained initial character feature vectors can be input into the aforementioned encoding layer to obtain the target character feature vector set. Finally, the target character feature vectors in the aforementioned target character feature vector set can be concatenated according to the character arrangement order to obtain the aforementioned initial user base vector.

[0080] As an example, the feature extraction sub-model described above could be a BiLSTM (Bidirectional Long Short-Term Memory) model. The feature annotation sub-model could be a CRF (Conditional Random Field) model. The word embedding layer could be an embedding layer. The encoding layer could be a Transformer layer.

[0081] The second sub-step involves inputting the initial user base vector into the feature extraction sub-model included in the initial user feature information extraction model to obtain the initial user feature vector.

[0082] The third sub-step involves inputting the initial user feature vector into the feature annotation sub-model included in the initial user feature information extraction model to obtain the initial user feature annotation vector.

[0083] The fourth sub-step involves determining the feature extraction loss value corresponding to the initial user feature annotation vector based on a preset loss function.

[0084] As an example, the preset loss function mentioned above can be, but is not limited to, one of the following: cross-entropy loss function, least squares function, or classification cross-entropy loss function.

[0085] The fifth sub-step is to determine the initial user feature information extraction model as the user feature information extraction model in response to the determination that the feature extraction loss value is less than the target threshold.

[0086] As an example, the target threshold mentioned above could be 0.01.

[0087] Optionally, in response to determining that the feature extraction loss value is greater than or equal to the target threshold, the aforementioned execution entity may also adjust the relevant parameters in the initial user feature information extraction model, determine the adjusted initial user feature information extraction model as the initial user feature information extraction model, and select a set of basic information of sample users from the aforementioned set of basic information of sample users, the basic information of each unselected sample user, and so on, for re-executing the aforementioned training sub-step. The relevant parameters in the initial user feature information extraction model can be adjusted using a preset adjustment algorithm.

[0088] As an example, the preset adjustment algorithm mentioned above can be the backpropagation algorithm or the stochastic gradient algorithm.

[0089] The relevant content of step 103, as an inventive point of this disclosure, solves the third technical problem mentioned in the background art: "wasting the time of the first user terminal to query practical learning information, resulting in a poor user experience." The factors that cause the wasted time of the first user terminal to query practical learning information and result in a poor user experience are often as follows: when sending a query to each associated second user terminal through the query server, the first user terminal needs to wait for a reply from each second user terminal, and the practical learning information replied by some second user terminals has low accuracy, thus reducing the accuracy of the obtained practical learning information. Solving the above factors can improve the accuracy of practical learning information. To achieve this effect, this disclosure can obtain the basic information of each associated second user terminal, and then convert the basic information into vectors through a vector transformation sub-model included in a pre-trained user feature information extraction model so that the model can perform feature extraction processing. Then, the feature information of the second user's basic information can be initially extracted through the feature extraction sub-model included in the pre-trained user feature information extraction model. Next, the feature information can be further extracted through the feature annotation sub-model included in the pre-trained user feature information extraction model so as to match the second user terminal and the query information, thereby determining that a reminder message can be sent. Finally, the second user terminal that receives the reminder information can generate practical learning information. Therefore, by extracting features from the basic information of the second user terminals, the second user terminals that need to receive reminder information can be determined. This eliminates the need to send reminder information to every single second user terminal, and the first user terminal can avoid waiting to receive practical learning information generated by second user terminals with low feature matching. This reduces the waste of communication resources, improves the accuracy of the obtained practical learning information, reduces the time wasted by the first user terminal in querying practical learning information, and enhances the user experience.

[0090] Step 104: Send the query result information sequence to the first user terminal.

[0091] In some embodiments, the executing entity may send the query result information sequence to the first user terminal. Thus, the first user terminal can receive the query result information sequence.

[0092] The above embodiments of this disclosure have the following beneficial effects: the practical learning information query method of some embodiments of this disclosure can reduce the wasted time of user terminals querying practical learning information and improve the user experience. Specifically, the reason for wasting user terminals' time querying practical learning information and resulting in a poor user experience is that in a structured database or file system, querying practical learning information only by searching for the name of the learning material is difficult because the name of the learning material cannot accurately summarize the content of the learning material, resulting in a decrease in the accuracy of the retrieved practical learning information, and the user terminal needs to repeatedly query practical learning information. Based on this, the practical learning information query method of some embodiments of this disclosure first obtains practical learning information query request information. Second, the above practical learning query request information is input into a pre-trained practical learning information query big model to obtain a query result information sequence. Thus, practical learning information can be initially retrieved through the big language model. Then, in response to determining that the above query result information sequence meets the preset sending conditions, each query result information sequence in the above query result information sequence is optimized to generate query summary information, resulting in a query summary information sequence. Thus, practical learning information that is easier for user terminals to understand can be obtained. Finally, the above query result information sequence is sent to the first user terminal. Therefore, the first user terminal can receive the queried practical learning information. Thus, some practical learning information query methods disclosed herein can query practical learning information from a knowledge base stored in the cloud or locally using a large model. Then, the queried information can be optimized to send the optimized practical learning information to the user terminal. This allows for the retrieval of learning materials from the knowledge base and the use of related learning materials as queried practical learning information. This improves the accuracy of the retrieved practical learning information, thereby reducing the number of times the user terminal repeatedly queries practical learning information, reducing wasted time on the user terminal, and improving the user experience.

[0093] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a practical learning information query device, which are similar to... Figure 1 Corresponding to the method embodiments shown, this practical learning information query device can be specifically applied to various electronic devices.

[0094] like Figure 2As shown, the practical learning information query device 200 in some embodiments includes: an acquisition unit 201, an input unit 202, an optimization unit 203, and a sending unit 204. The acquisition unit 201 is configured to acquire practical learning information query request information. The input unit 202 is configured to input the practical learning query request information into a practical learning information query big model to obtain a query result information sequence. The optimization unit 203 is configured to, in response to determining that the query result information sequence meets preset sending conditions, optimize each query result information sequence in the query result information sequence to generate query summary information, obtaining a query summary information sequence. The sending unit 204 is configured to send the query result information sequence to a first user terminal.

[0095] It is understandable that the units recorded in the practical learning information query device 200 are related to the reference. Figure 1 The steps in the described practical learning information retrieval method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the practical learning information retrieval method also apply to the practical learning information retrieval device 200 and the units contained therein, and will not be repeated here.

[0096] The following is for reference. Figure 3 This document illustrates a structural schematic of an electronic device 300 suitable for implementing some embodiments of the present disclosure. The electronic devices in some embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 3 The terminal device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0097] like Figure 3 As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0098] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0099] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0100] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0101] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0102] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: acquire practical learning information query request information; input the practical learning query request information into a practical learning information query big model to obtain a query result information sequence; in response to determining that the query result information sequence meets preset sending conditions, optimize each query result information sequence in the query result information sequence to generate query summary information, obtaining a query summary information sequence; and send the query result information sequence to a first user terminal.

[0103] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0104] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0105] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including an acquisition unit, an input unit, an optimization unit, and a transmission unit. The names of these units do not necessarily limit the specific unit; for example, an acquisition unit may also be described as a "unit for acquiring practical learning information query request information."

[0106] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0107] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for retrieving practical learning information, comprising: Request information for obtaining practical learning information; The practical learning information query request information is input into the practical learning information query model to obtain a sequence of query result information, including: Acquire a knowledge base for practical learning; Each practical learning knowledge base in the practical learning knowledge base set is transformed to generate a practical learning knowledge graph, resulting in a practical learning knowledge graph set. Based on the practical learning query request information, the practical learning knowledge base set is subjected to a first matching process to obtain a first matching knowledge information set; Based on the practice learning knowledge graph set and the practice learning query request information, the first matching knowledge information set is filtered to obtain a filtered knowledge information set. This includes: vectorizing the practice learning query request information to obtain a query request vector; for each first matching knowledge information in the first matching knowledge information set, the following transformation steps are performed: determining the practice learning knowledge graph in the practice learning knowledge graph set corresponding to the first matching knowledge information as the target knowledge graph; concatenating each target knowledge vector in the target knowledge vector set according to the order of characters in the first matching knowledge information to obtain a target knowledge concatenation vector; determining the query similarity value corresponding to the query request vector and the target knowledge concatenation vector; sorting the determined query similarity values ​​in descending order to obtain a query similarity value sequence; and determining the first matching knowledge information corresponding to the first target number of query similarity values ​​in the query similarity value sequence as filtered knowledge information to obtain the filtered knowledge information set. Based on the practical learning query request information, the filtered knowledge information set is subjected to a second matching process to obtain a second matched knowledge information set. The practical learning query request information and the second matching knowledge information set are fused to obtain model query instruction information, wherein the practical learning query request information and the second matching knowledge information set are determined as the practical learning query request information and the second matching knowledge information set included in the model query instruction information; The model query instruction information is input into the large-scale practical learning information query model to obtain the query result information sequence; In response to determining that the query result information sequence meets the preset sending conditions, each query result information sequence in the query result information sequence is optimized to generate query summary information, thus obtaining a query summary information sequence; The query result information sequence is sent to the first user terminal.

2. The method according to claim 1, wherein, The method further includes: In response to determining that the query result information sequence does not meet the preset sending conditions, a practical learning reminder message is generated based on the practical learning information query request information; In response to receiving a reminder message sending instruction from the first user terminal, the practical learning reminder message is sent to the practical learning server and each of the second user terminals corresponding to the practical learning reminder message; The system receives practical learning information from each of the second user terminals to obtain a practical learning information set, and then sends the practical learning information set to the first user terminal.

3. The method according to claim 2, wherein, The step of generating practical learning reminder information based on the practical learning information query request information includes: Obtain the second user's basic information set; Each piece of second user basic information in the second user basic information set is input into a pre-trained user feature information extraction model to generate second user feature information, thus obtaining the second user feature information set. The practice learning information query request information and each second user feature information in the second user feature information set are matched to generate second user matching information, thus obtaining the second user matching information set. The second user basic information corresponding to the second user matching information in the second user matching information set that meets the preset matching conditions is determined as the target second user basic information, and the target second user basic information set is obtained. The target second user basic information set and the practical learning information query request information are fused together to obtain the practical learning reminder information.

4. A practical learning information retrieval device, comprising: The acquisition unit is configured to retrieve practical learning information query requests. The input unit is configured to input the practical learning information query request information into the practical learning information query big model to obtain a query result information sequence; The input unit is further configured as follows: Acquire a knowledge base for practical learning; Each practical learning knowledge base in the practical learning knowledge base set is transformed to generate a practical learning knowledge graph, resulting in a practical learning knowledge graph set. Based on the practical learning query request information, the practical learning knowledge base set is subjected to a first matching process to obtain a first matching knowledge information set; Based on the practice learning knowledge graph set and the practice learning query request information, the first matching knowledge information set is filtered to obtain a filtered knowledge information set. This includes: vectorizing the practice learning query request information to obtain a query request vector; for each first matching knowledge information in the first matching knowledge information set, the following transformation steps are performed: determining the practice learning knowledge graph in the practice learning knowledge graph set corresponding to the first matching knowledge information as the target knowledge graph; concatenating each target knowledge vector in the target knowledge vector set according to the order of characters in the first matching knowledge information to obtain a target knowledge concatenation vector; determining the query similarity value corresponding to the query request vector and the target knowledge concatenation vector; sorting the determined query similarity values ​​in descending order to obtain a query similarity value sequence; and determining the first matching knowledge information corresponding to the first target number of query similarity values ​​in the query similarity value sequence as filtered knowledge information to obtain the filtered knowledge information set. Based on the practical learning query request information, the filtered knowledge information set is subjected to a second matching process to obtain a second matched knowledge information set. The practical learning query request information and the second matching knowledge information set are fused to obtain model query instruction information, wherein the practical learning query request information and the second matching knowledge information set are determined as the practical learning query request information and the second matching knowledge information set included in the model query instruction information; The model query instruction information is input into the large-scale practical learning information query model to obtain the query result information sequence; An optimization unit is configured to optimize each query result information sequence in the query result information sequence to generate query summary information in response to determining that the query result information sequence meets a preset sending condition, thereby obtaining a query summary information sequence. The sending unit is configured to send the query result information sequence to the first user terminal.

5. An electronic device, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-3.

6. A computer-readable medium having a computer program stored thereon, wherein, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-3.

Images