Agricultural knowledge recommendation method and device

Abstract

The present invention provides an agricultural knowledge recommendation method and device, the method comprising: inputting the knowledge acquisition request information and attribute information of the target user into an agricultural knowledge recommendation model, obtaining the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model; and determining the target recommended knowledge from an agricultural knowledge data set according to the target recommended knowledge category. The agricultural knowledge recommendation method of the present invention determines the attribute information of the target user, uses the agricultural knowledge recommendation model to determine the accurate target recommended knowledge category, and determines the target recommended knowledge from the constructed agricultural knowledge data set to push to the target user. It can accurately determine the push information based on the attribute information of the target user, so that the knowledge platform can carry out precise knowledge recommendation services to push precise agricultural knowledge to different users.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural knowledge services, and in particular to an agricultural knowledge recommendation method and device. Background Art

[0002] In the field of agricultural knowledge services, with the continuous expansion of the service chain, the increasing number of user types, the continuous expansion of the user scale, and the continuous in-depth integration of modern information technology and agricultural development, the categories of agricultural knowledge are constantly enriched and the quantity is constantly accumulated. Various types of knowledge platforms have basically formed a massive knowledge base with wide field coverage and complete knowledge classification, providing a massive amount of agricultural knowledge for agricultural production and management practitioners. Agricultural knowledge services have crossed the contradiction between user information retrieval needs and data scarcity in the early stage of Internet platform development, and entered a new stage of contradiction between large-scale redundant data and user needs for accurate information acquisition.

[0003] However, in the process of knowledge service, there is little research and application on user interest discovery and accurate knowledge recommendation, which makes users face a huge amount of knowledge, resulting in the problem of "knowledge overload" where redundant knowledge interferes with the accurate positioning of useful knowledge, and the problem of "knowledge disorientation" where users with low cultural level and information acquisition skills are easily lost in the ocean of knowledge and find it difficult to find the required knowledge. How to accurately push appropriate agricultural knowledge to different users is an urgent problem to be solved. Summary of the invention

[0004] The present invention provides an agricultural knowledge recommendation method and device, which are used to solve the defect that it is difficult to push accurate agricultural knowledge to users in the prior art, and to achieve accurate knowledge push according to the needs of different users.

[0005] The present invention provides an agricultural knowledge recommendation method, comprising:

[0006] Determine knowledge acquisition request information of a target user and attribute information of the target user;

[0007] Inputting the knowledge acquisition request information and attribute information of the target user into the agricultural knowledge recommendation model to obtain the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model;

[0008] According to the target recommended knowledge category, determining target recommended knowledge from the agricultural knowledge data set;

[0009] Pushing the target feedback knowledge and the target recommendation knowledge to the target user;

[0010] Among them, the agricultural knowledge recommendation model is trained with historical knowledge acquisition request information of the knowledge platform and attribute information of users corresponding to the historical knowledge acquisition request information as samples, and feedback knowledge and recommended knowledge categories corresponding to the historical knowledge acquisition request information as labels.

[0011] According to an agricultural knowledge recommendation method provided by the present invention, the agricultural knowledge dataset is determined by:

[0012] Acquire the attribute information of the user corresponding to each historical knowledge acquisition request information, the feedback knowledge corresponding to each historical knowledge acquisition request information, and the recommended knowledge category;

[0013] Based on the feedback knowledge corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information is determined; the feedback knowledge corresponding to each historical knowledge acquisition request information is determined based on the knowledge graph information associated with each historical knowledge acquisition request information;

[0014] Mapping the user's attribute information corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information, and the recommended knowledge category corresponding to each historical knowledge acquisition request information into feature vectors, and combining the feature vectors corresponding to each historical knowledge acquisition request information into a combined feature vector in a preset order;

[0015] Based on the combined feature vectors corresponding to each historical knowledge acquisition request information, the agricultural knowledge data set is obtained.

[0016] According to an agricultural knowledge recommendation method provided by the present invention, the user's attribute information includes portrait attribute information and behavior attribute information; the portrait attribute information includes at least one of the user's region, industry, related agricultural crop varieties, gender, age, and education level; the behavior attribute information includes at least one of the time when the user clicks on the knowledge on the knowledge platform, the knowledge category clicked, the reading stay time, whether there are comments, whether the user asks the question himself, and whether the question is adopted;

[0017] Mapping the user's attribute information corresponding to each historical knowledge acquisition request information into a feature vector includes:

[0018] Mapping each attribute information in the portrait attribute information and the behavior attribute information corresponding to each historical knowledge acquisition request information into an attribute sub-vector respectively;

[0019] All the attribute sub-vectors corresponding to the portrait attribute information and the behavior attribute information corresponding to each historical knowledge acquisition request information are combined to obtain a feature vector of the attribute information of each user corresponding to each historical knowledge acquisition request information.

[0020] According to an agricultural knowledge recommendation method provided by the present invention, the knowledge graph information associated with the historical knowledge acquisition request information includes at least one of the category of candidate knowledge, knowledge representation form, knowledge source, publishing location, publishing time, number of comments, and whether it has been adopted;

[0021] Mapping the knowledge graph information associated with each historical knowledge acquisition request information into a feature vector includes:

[0022] Mapping each type of information in the knowledge graph information associated with each historical knowledge acquisition request information into a knowledge graph sub-vector;

[0023] All types of knowledge graph sub-vectors in the knowledge graph information associated with each historical knowledge acquisition request information are combined to obtain the feature vectors of each knowledge graph information associated with each historical knowledge acquisition request information.

[0024] According to an agricultural knowledge recommendation method provided by the present invention, the agricultural knowledge recommendation model is obtained by training based on a wide and deep interest network model, and the agricultural knowledge recommendation model includes an input layer, a class hidden layer and an output layer;

[0025] The input layer is used to receive the knowledge acquisition request information and attribute information of the target user;

[0026] The class hidden layer is used to obtain the target feedback knowledge and the target recommended knowledge category;

[0027] The output layer is used to output the target feedback knowledge and the target recommended knowledge category.

[0028] According to an agricultural knowledge recommendation method provided by the present invention, the step of determining the knowledge acquisition request information of a target user and the attribute information of the target user comprises:

[0029] receiving a target input from the target user;

[0030] Based on the target input, determine the knowledge acquisition request information of the target user, and obtain the historical usage data of the target user on the knowledge platform;

[0031] Based on the registration information, current usage data and historical usage data of the target user on the knowledge platform, attribute information of the target user is determined.

[0032] The present invention also provides an agricultural knowledge recommendation device, comprising:

[0033] A first processing module, used to determine knowledge acquisition request information of a target user and attribute information of the target user;

[0034] The second processing module is used to input the knowledge acquisition request information and attribute information of the target user into the agricultural knowledge recommendation model to obtain the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model;

[0035] A third processing module is used to determine target recommended knowledge from the agricultural knowledge data set according to the target recommended knowledge category;

[0036] A fourth processing module, used for pushing the target feedback knowledge and the target recommendation knowledge to the target user;

[0037] Among them, the agricultural knowledge recommendation model is trained with historical knowledge acquisition request information of the knowledge platform and attribute information of users corresponding to the historical knowledge acquisition request information as samples, and feedback knowledge and recommended knowledge categories corresponding to the historical knowledge acquisition request information as labels.

[0038] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the agricultural knowledge recommendation methods described above is implemented.

[0039] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the agricultural knowledge recommendation method as described in any one of the above is implemented.

[0040] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the agricultural knowledge recommendation method as described above is implemented.

[0041] The agricultural knowledge recommendation method and device provided by the present invention determine the attribute information of the target user, use the agricultural knowledge recommendation model to determine the accurate target recommended knowledge category, and determine the target recommended knowledge from the constructed agricultural knowledge data set to push to the target user. It can accurately determine the push information based on the attribute information of the target user, so that the knowledge platform can carry out precise knowledge recommendation services to push precise agricultural knowledge to different users. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1This is one of the flow charts of the agricultural knowledge recommendation method provided by the present invention;

[0044] Figure 2 It is a schematic diagram of the generation process of the combined feature vector provided by the present invention;

[0045] Figure 3 This is the second flow chart of the agricultural knowledge recommendation method provided by the present invention;

[0046] Figure 4 This is the third flow chart of the agricultural knowledge recommendation method provided by the present invention;

[0047] Figure 5 It is a structural schematic diagram of the agricultural knowledge recommendation device provided by the present invention;

[0048] Figure 6 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Combine the following Figure 1-Figure 6 The agricultural knowledge recommendation method and device of the present invention are described.

[0051] The execution subject of the agricultural knowledge recommendation method of the embodiment of the present invention may be a processor. Of course, in some embodiments, the execution subject may also be a server. The specific type of the execution subject is not limited here. The following takes the execution subject as an example to illustrate the agricultural knowledge recommendation method of the embodiment of the present invention.

[0052] like Figure 1 As shown, the agricultural knowledge recommendation method of the embodiment of the present invention mainly includes step 110, step 120, step 130 and step 140.

[0053] Step 110: determine the knowledge acquisition request information of the target user and the attribute information of the target user.

[0054] It is understandable that the target users are users of the knowledge platform, and the target users can obtain agriculture-related knowledge through the knowledge platform.

[0055] The target user's knowledge acquisition request information can be the target user's search information for the required knowledge on the knowledge platform, or it can be the question information published on the knowledge platform. In other words, the target user can obtain the knowledge he needs through the knowledge acquisition request information.

[0056] It is understandable that the target user's attribute information may be information associated with the target user on the knowledge platform. For example, the target user's attribute information includes image attribute information and behavior attribute information.

[0057] The portrait attribute information is the identity information of the target user itself, and the portrait attribute information includes at least one of the user's region, industry, related agricultural crop varieties, gender, age, and education level.

[0058] The behavior attribute information is the behavior characteristic information of the target user on the knowledge platform, and the behavior attribute information includes at least one of the time when the user clicks on the knowledge on the knowledge platform, the knowledge category clicked, the reading dwell time, whether there are comments, whether the user asks the question himself, and whether the answer is adopted.

[0059] In some embodiments, step 110: determining the knowledge acquisition request information of the target user and the attribute information of the target user may include the following steps.

[0060] In this implementation, a target input of a target user may be received.

[0061] It is understandable that the target users are generally groups such as farmers, and the target user groups usually use mobile terminals such as mobile phones, tablets, etc. to achieve target input.

[0062] On this basis, taking a mobile phone as an example, target input can be expressed in at least one of the following ways:

[0063] First, the target input may be a touch operation, including but not limited to a click operation and a press operation.

[0064] In this implementation, receiving the target input of the target user may be performed by receiving a touch operation of the user on the display area of ​​the knowledge platform related module on the mobile phone display screen.

[0065] For example, a touch input of a target user clicking on a relevant agricultural question may be received, or an agricultural-related question input by the target user may be received, thereby determining the knowledge acquisition request information of the target user.

[0066] Second, the target input may be in the form of voice input.

[0067] In this implementation, the target voice may trigger the mobile phone to record the target user's knowledge acquisition request information.

[0068] For example, users can use their mobile phones to directly input their agriculture-related questions into the knowledge platform by voice.

[0069] Third, the target input can be expressed as physical key input.

[0070] In this embodiment, input can be achieved through a physical keyboard or the like.

[0071] Of course, in other embodiments, the target input may also be in other forms, including but not limited to character input, etc., which can be determined according to actual needs and is not limited in this embodiment of the present application.

[0072] On this basis, the target user's knowledge acquisition request information is determined based on the target input.

[0073] It is understandable that the historical usage data of the target user on the knowledge platform can be obtained based on the target input of the target user.

[0074] Furthermore, the attribute information of the target user may be determined based on the registration information, current usage data, and historical usage data of the target user on the knowledge platform.

[0075] It is understandable that when users register on the knowledge platform, they will provide information such as their basic identity information, location information, and preferred content areas.

[0076] In the process of historical and current use of the knowledge platform, users will generate a large amount of usage data, which can be collected and analyzed to obtain attribute information of the target users.

[0077] In this embodiment, by collecting and analyzing the user's usage data and registration data, the attribute information of each target user can be obtained, and then the knowledge needs of each target user can be obtained more accurately.

[0078] Step 120, inputting the knowledge acquisition request information and attribute information of the target user into the agricultural knowledge recommendation model, and obtaining the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model.

[0079] It can be understood that the agricultural knowledge recommendation model can be trained based on the wide and deep interest network model.

[0080] In the embodiment of the present invention, a wide and deep interest network (WDIN) model for agricultural knowledge recommendation can be constructed, and an interest network model combining a wide neural network and a deep neural network for predicting user interests using a neuromorphic calculation mechanism (Neuromorphological calculation, NeuroC) is introduced to capture the user's specific interest points, which is used to represent the user's preference interest in a certain type of knowledge. The wide and deep interest network model network can be divided into an input layer, a class hidden layer, and an output layer.

[0081] The input layer is used to receive the target user's knowledge acquisition request information and attribute information. The input layer includes neuron nodes and brain-like neuron presynaptic nodes. The neuron nodes in the input layer have the same functions as the traditional deep neural network neuron nodes and are fully connected with the next layer; the brain-like neuron presynaptic nodes are directly connected to the brain-like neurons to enhance the efficiency of information transmission, reduce the probability of information loss, and avoid increasing the model complexity and losing important information after unnecessary hidden layer calculations.

[0082] The class hidden layer is used to obtain target feedback knowledge and target recommendation knowledge categories. The class hidden layer is similar to the hidden layer of a conventional neural network, but each layer adds a class of brain neuron nodes. The brain-like neuron nodes do not participate in the parameter transfer of the previous layer, but only transfer parameters to the next layer. The activation function of the brain-like neuron nodes in each layer follows the principle that the closer to the output layer, the lower the activation function order. Starting from simulating the five major perception systems of human vision, hearing, touch, smell, and taste, the class hidden layer is used to simulate the perception system. Combined with the principles of balancing the width and depth of the model and reducing the parameter scale, 5 class hidden layers are selected.

[0083] The output layer is used to output the target feedback knowledge and the target recommended knowledge category, that is, the output layer can derive the specific recommended knowledge category.

[0084] The training of the agricultural knowledge recommendation model involving brain-like neurons in the present invention is divided into the following stages, namely, establishing connections, building networks, and selecting activation functions.

[0085] 1. Establish a connection.

[0086] The initial network structure uses omnidirectional connection, the initial number of nodes in the input layer is 1, and the number of nodes in the hidden layer is determined according to the complexity of the actual problem being processed. Each hidden layer sets a brain-like neuron as a foreign key node, and n groups of brain-like neuron presynaptic nodes are established in the input layer. The number of brain-like neuron presynaptic node groups is the same as the number of hidden layers. Each group corresponds to a brain-like neuron node in the hidden layer, and the connection relationship between each brain-like neuron presynaptic node and the brain-like neuron is recorded. The number of brain-like neuron presynaptic nodes in each group is not fixed and is determined according to the actual activation of the network simulation.

[0087] 2. Network construction.

[0088] The brain-like neuron presynaptic nodes and brain-like neurons are involved in network calculations. The synapses are event-driven. When the activation conditions of the brain-like neuron presynaptic nodes are met, there will be pulses, and the brain-like neurons will perform synaptic updates. If a group of brain-like neuron presynaptic nodes is activated, a new node, the brain-like neuron presynaptic node, will be added to the group for subsequent parameter use.

[0089] Activation conditions for the presynaptic nodes of brain-like neurons: Using the horizontal optimization method, each independent variable is traversed through each input layer node and presynaptic node of brain-like neurons in turn, and the node with the highest fitting accuracy is selected as the input node of the variable; until each independent variable finds the corresponding input layer node or presynaptic node of brain-like neurons with the highest fitting accuracy.

[0090] 3. Activation function selection.

[0091] The agricultural knowledge recommendation model of the embodiment of the present invention is based on the pulse time dependent plasticity (STDP) mechanism of synaptic connection change between neurons, and proposes an adaptive activation function of brain-like neurons to assist in model training. In terms of the activation function of brain-like neurons, considering the neuron conduction law and model interpretability, an approximate Taylor polynomial of a common smooth function (Taylor polynomial with residual terms removed) is selected as the benchmark activation function. Taylor polynomials can represent a function with a number of consecutive additions. These added terms are obtained by the derivative of the function at a certain point, which can fit the function characteristics similar to neuron conduction, facilitating the interpretation of model laws.

[0092] The benchmark activation function is used as the brain-like neuron activation function of the first hidden layer, and the first-order derivative, second-order derivative, third-order derivative, and fourth-order derivative of the benchmark activation function are used as the brain-like neuron activation functions of the second, third, fourth, and fifth hidden layers, respectively.

[0093] The alternative smoothing function must meet the following conditions:

[0094] In the domain of definition, the function is monotonic and a strictly increasing function, that is, in the interval x∈[0,1], for any two values ​​of the independent variables x1 and x2, when x1>x2, f(x1)≥f(x2), then f(x) is said to be an increasing function in this interval (another way of saying it is a monotonic non-decreasing function).

[0095] exist x ∈[0,1], the function is uniformly differentiable, that is, suppose the function F is defined on [0,1]. If there is a function f and a positive number M defined on [0,1], and a non-negative function d(x) that increases on [0,1], and is unbounded on [0,1], so that for anyx and x+h, we have the following inequality: |F(x+h)-F(x)-f(x)h|≤|hd(|h|)|;

[0096] In this case, F is said to be uniformly differentiable on [0,1], and f(x) is said to be the derivative of F(x), denoted by F'(x) = f(x). The above inequality can also be written in the equivalent form of an equation: F(x+h)-F(x) = f(x)h+M(x,h)hd(|h|);

[0097] Here, M(x,h) is a bounded function on the region {(x,h):x∈[a,b],x+h∈[a,b]}.

[0098] Based on the above constraints, choose e x For a smoothing function, select e x The approximate Taylor polynomial of is used as the benchmark activation function, that is, the brain-like neuron activation function of the first hidden layer.

[0099] Among them, e x The Taylor polynomial of

[0100] e x The approximate Taylor polynomial of is

[0101] Let F1(x) be the brain-like neuron activation function of the first hidden layer, denoted as:

[0102]

[0103] Let F2(x) be the activation function of the brain-like neuron in the third hidden layer, denoted as:

[0104]

[0105] Let F3(x) be the activation function of the brain-like neuron in the third hidden layer, denoted as:

[0106]

[0107] set up F4 (x) The brain-like neuron activation function of the fourth hidden layer is expressed as:

[0108]

[0109] Let F5(x) be the activation function of the brain-like neuron in the fifth hidden layer, denoted as:

[0110]

[0111] After the construction of the agricultural knowledge recommendation model and the determination of the initial parameters are completed, the training data is input into the agricultural knowledge recommendation model. In some embodiments, the agricultural knowledge recommendation model is trained using the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and the feedback knowledge corresponding to the historical knowledge acquisition request information and the recommended knowledge category as labels.

[0112] In this case, after the target user's knowledge acquisition request information and attribute information are input into the agricultural knowledge recommendation model, the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model can be obtained.

[0113] In this embodiment, by proposing a wide and deep interest network model, combining the advantages of wide neural networks, deep neural networks and neuron computing, improving the network activation mechanism and activation function, and introducing a neuromorphic computing mechanism to sensitively distinguish differences in user attribute information, the shortcomings of single knowledge recommendation methods such as knowledge graphs and deep learning cannot fully cover user knowledge needs, cannot accurately match user demand knowledge, and cannot flexibly track changes in user interests are overcome. The accuracy of user knowledge demand prediction is improved, and the user interest changes have a more sensitive perception ability, which can realize accurate and timely knowledge recommendation services.

[0114] In addition, the agricultural knowledge recommendation method proposed in the present invention also has the following advantages.

[0115] The agricultural knowledge recommendation model of the embodiment of the present invention is easier to understand: the model is highly interpretable and can realize the reverse deduction of the user's knowledge selection rules. The shorter the key distance between the variables in the input layer and the output layer, the more obvious the action mechanism is, the stronger the correlation is, and the greater the impact of adjusting this variable indicator on the final result.

[0116] The prediction accuracy of the model in the embodiment of the present invention is improved. The brain-like neurons only transfer parameters with the lower layer, which can reduce the influence weight of weakly correlated independent variables, effectively solve the influence of weakly correlated indicators on the output results, reduce the occurrence of overfitting, and more easily obtain the global optimal solution;

[0117] The parameter scale of the embodiment of the present invention is reduced. Under the parameter scale of controllable width and depth, the agricultural knowledge recommendation model of the embodiment of the present invention achieves a good prediction effect, reduces the "gradient vanishing" effect of the traditional model, and avoids the problem that the lower layer basically cannot receive effective training signals after the gradient exponential decay, resulting in slow training efficiency and low model accuracy. In addition, since the independent variables are distributed in the input layer neuron nodes and the presynaptic nodes of the brain-like neurons according to the optimization scheme, the network parameter scale becomes smaller, which greatly eliminates the problem of parameter quantity expansion.

[0118] Step 130 , determining target recommended knowledge from the agricultural knowledge dataset according to the target recommended knowledge category.

[0119] It can be understood that after determining the target recommended knowledge category, the knowledge corresponding to the target recommended knowledge category can be further determined.

[0120] In some embodiments, the knowledge corresponding to the target recommended knowledge category in the relevant knowledge graph can be directly obtained.

[0121] In this embodiment, in order to further ensure the relevance between the knowledge corresponding to the target recommended knowledge category and the target user, the target recommended knowledge can be determined from the agricultural knowledge data set.

[0122] It should be noted that the agricultural knowledge dataset can store data corresponding to user attribute information, relevant knowledge graphs, and recommended knowledge categories, so as to determine knowledge that better suits user needs based on user attribute information.

[0123] In some embodiments, the agricultural knowledge dataset is determined in the following manner.

[0124] In this embodiment, the attribute information of the user corresponding to each historical knowledge acquisition request information, the feedback knowledge corresponding to each historical knowledge acquisition request information, and the recommended knowledge category may be acquired.

[0125] In this case, the historical data of the user in the knowledge platform is identified and analyzed. The feedback knowledge corresponding to each historical knowledge acquisition request information may be the answer information of the user's historical questions.

[0126] It can be understood that the knowledge graph information associated with each historical knowledge acquisition request information can be determined based on the feedback knowledge corresponding to each historical knowledge acquisition request information, that is, the knowledge graph information corresponding to the historical answer information can be determined through the historical answer information to determine the possible knowledge source of the historical answer information.

[0127] In this embodiment, the feedback knowledge corresponding to each historical knowledge acquisition request information is determined based on the knowledge graph information associated with each historical knowledge acquisition request information, and the source of the historical answer information is the corresponding knowledge graph information.

[0128] In this case, the attribute information of the user corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information, and the recommended knowledge category corresponding to each historical knowledge acquisition request information are all mapped into feature vectors, and the feature vectors corresponding to each historical knowledge acquisition request information are combined into a combined feature vector in a preset order so that the agricultural knowledge recommendation model can recognize and process it.

[0129] For example, the feature vector corresponding to the user's attribute information is A, the feature vector corresponding to the knowledge graph information associated with each historical knowledge acquisition request information is B, and the feature vector of the recommended knowledge category corresponding to each historical knowledge acquisition request information is C, then the combined feature vector can be [A, B, C].

[0130] It can be understood that, by obtaining the combined feature vectors corresponding to the request information according to different historical knowledge, multiple groups of combined feature vectors can be obtained, and then the agricultural knowledge data set can be obtained.

[0131] In some embodiments, the user's attribute information includes portrait attribute information and behavioral attribute information; the portrait attribute information includes at least one of the user's region, industry, related agricultural crop varieties, gender, age, and educational background; the behavioral attribute information includes at least one of the time when the user clicks on knowledge on the knowledge platform, the category of knowledge clicked, the reading dwell time, whether there are any comments, whether the question was asked by the user himself, and whether the answer was adopted.

[0132] Mapping the user's attribute information corresponding to each historical knowledge acquisition request information into a feature vector includes the following process.

[0133] In this embodiment, each attribute information in the portrait attribute information and the behavior attribute information corresponding to each historical knowledge acquisition request information is mapped into an attribute sub-vector respectively.

[0134] The attribute sub-vector corresponding to the portrait attribute information corresponding to each historical knowledge acquisition request information may be A1, and the attribute sub-vector corresponding to the behavior attribute information corresponding to each historical knowledge acquisition request information may be A2.

[0135] On this basis, all attribute sub-vectors corresponding to the portrait attribute information and the behavior attribute information corresponding to each historical knowledge acquisition request information are combined to obtain the feature vector of the attribute information of each user corresponding to each historical knowledge acquisition request information.

[0136] For example, the user's region, industry, related agricultural crop varieties, gender, age, and education level can be mapped to vectors of the same length A11, A12, A13, A14, A15, and A16. The time when the user clicks on knowledge on the knowledge platform, the knowledge category clicked, the reading stay time, whether there are comments, whether the user asks the question himself, and whether the question is adopted can be mapped to vectors of the same length A21, A22, A23, A24, A25, and A26.

[0137] Therefore, A1 can be expressed as [A11, A12, A13, A14, A15, A16], and A2 can be expressed as [A21, A22, A23, A24, A25, A26]. On this basis, A can be expressed as [A1, A2], and the combined feature vector can be [A1, A2, B, C].

[0138] like Figure 2 As shown, the user's portrait attribute information, behavior attribute information, and knowledge graph information are mapped to the interest space domain according to the corresponding knowledge categories to generate a user context mixed sequence vector, that is, the combined feature vector can be [A1, A2, B, C].

[0139] like Figure 3 As shown, in this embodiment, an interest space domain representing the characteristic attributes of different types of users can be constructed based on the historical data and new data of the knowledge platform, and the user's portrait attribute information, behavioral attribute information, and knowledge graph information can be mapped to the interest space domain according to the corresponding knowledge category classification to form a vector combination, combined with the vector of the recommended knowledge category, and an agricultural knowledge data set is formed through vector mapping, so that the agricultural knowledge recommendation model can use deep neural networks, wide neural networks, and neural state computing methods to output the recommended knowledge categories required by users.

[0140] In the prior art, in the field of agricultural knowledge services, user contexts include user portrait features, user behavior sequences, and knowledge graph features. Previous deep interest prediction models performed embedding mapping on features without considering user feature differences, and violently mapped the features of the input layer into a vector of fixed length, resulting in the user's interest features being compressed into an embedding vector of fixed length regardless of the implicit user interest. This method makes it difficult to capture the user's diverse interest differences in terms of region, industry, etc.

[0141] In order to solve the difficulties in matching user's interested knowledge from rich historical knowledge, the present invention fully considers the characteristics of complex agricultural knowledge service domain scenarios, changeable user interests, unclear demand expression, etc. in the process of knowledge demand prediction. By integrating user portrait features, user behavior features, knowledge graph features and other user knowledge demand influencing factors, the present invention comprehensively analyzes the industry differences, habit differences, and interest changes with farming seasons of users in the agricultural field, and constructs a user context mixed sequence for agricultural knowledge recommendation, thereby improving the accuracy of user knowledge demand prediction, having a more sensitive perception ability to user interest changes, and realizing accurate and timely knowledge recommendation services.

[0142] Step 140: Push the target feedback knowledge and the target recommendation knowledge to the target user.

[0143] On this basis, the target feedback knowledge and target recommendation knowledge are pushed to the target users so that the agricultural knowledge platform can carry out precise knowledge recommendation services.

[0144] like Figure 4 As shown, the agricultural knowledge extraction module of the agricultural knowledge recommendation model in the agricultural knowledge recommendation system of the knowledge platform can be used to extract relevant knowledge, and the knowledge push module and the user's terminal service can be used to push knowledge to the user.

[0145] It is understandable that historical information can be pushed based on the historical data of the target user to achieve interest-tracing push, or targeted push of relevant regional industries can be performed based on the location of the target user, or scheduled push can be performed based on the target user's historical login time.

[0146] According to the agricultural knowledge recommendation method provided by the embodiment of the present invention, by determining the attribute information of the target user, the agricultural knowledge recommendation model is used to determine the accurate target recommended knowledge category, and the target recommended knowledge is determined from the constructed agricultural knowledge data set to be pushed to the target user. The push information can be accurately determined based on the attribute information of the target user, so that the knowledge platform can carry out precise knowledge recommendation services to push precise agricultural knowledge to different users.

[0147] The agricultural knowledge recommendation device provided by the present invention is described below. The agricultural knowledge recommendation device described below and the agricultural knowledge recommendation method described above can be referenced to each other.

[0148] like Figure 5 As shown, the agricultural knowledge recommendation device according to the embodiment of the present invention mainly includes a first processing module 510 , a second processing module 520 , a third processing module 530 and a fourth processing module 540 .

[0149] The first processing module 510 is used to determine the knowledge acquisition request information of the target user and the attribute information of the target user;

[0150] The second processing module 520 is used to input the knowledge acquisition request information and attribute information of the target user into the agricultural knowledge recommendation model, and obtain the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model;

[0151] The third processing module 530 is used to determine target recommended knowledge from the agricultural knowledge data set according to the target recommended knowledge category;

[0152] The fourth processing module 540 is used to push the target feedback knowledge and the target recommendation knowledge to the target user;

[0153] Among them, the agricultural knowledge recommendation model is trained with the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and the feedback knowledge corresponding to the historical knowledge acquisition request information and the recommended knowledge category as labels.

[0154] According to the agricultural knowledge recommendation device provided by the embodiment of the present invention, by determining the attribute information of the target user, using the agricultural knowledge recommendation model to determine the accurate target recommended knowledge category, and determining the target recommended knowledge from the constructed agricultural knowledge data set to push to the target user, it is possible to accurately determine the push information based on the attribute information of the target user, so as to facilitate the knowledge platform to carry out precise knowledge recommendation services, so as to push precise agricultural knowledge to different users.

[0155] In some embodiments, the third processing module 530 is also used to obtain the attribute information of the user corresponding to each historical knowledge acquisition request information, the feedback knowledge corresponding to each historical knowledge acquisition request information, and the recommended knowledge category; based on the feedback knowledge corresponding to each historical knowledge acquisition request information, determine the knowledge graph information associated with each historical knowledge acquisition request information; the feedback knowledge corresponding to each historical knowledge acquisition request information is determined based on the knowledge graph information associated with each historical knowledge acquisition request information; the attribute information of the user corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information, and the recommended knowledge category corresponding to each historical knowledge acquisition request information are all mapped into feature vectors, and the feature vectors corresponding to each historical knowledge acquisition request information are combined into a combined feature vector in a preset order; based on the combined feature vector corresponding to each historical knowledge acquisition request information, an agricultural knowledge data set is obtained.

[0156] In some embodiments, the user's attribute information includes portrait attribute information and behavioral attribute information; the portrait attribute information includes at least one of the user's region, industry, related agricultural crop varieties, gender, age, and education level; the behavioral attribute information includes at least one of the time when the user clicks on the knowledge on the knowledge platform, the category of knowledge clicked, the reading dwell time, whether there are comments, whether the user asked the question himself, and whether the answer was adopted; the third processing module 530 is also used to map the portrait attribute information corresponding to each historical knowledge acquisition request information and each attribute information in the behavioral attribute information into attribute sub-vectors respectively; all attribute sub-vectors corresponding to the portrait attribute information and the behavioral attribute information corresponding to each historical knowledge acquisition request information are combined to obtain the feature vector of the attribute information of each user corresponding to each historical knowledge acquisition request information.

[0157] In some embodiments, the knowledge graph information associated with the historical knowledge acquisition request information includes at least one of the category of candidate knowledge, knowledge representation form, knowledge source, publication location, publication time, number of comments, and whether it has been adopted; the third processing module 530 is also used to map each type of information in the knowledge graph information associated with each historical knowledge acquisition request information into a knowledge graph sub-vector; combine all types of knowledge graph sub-vectors in the knowledge graph information associated with each historical knowledge acquisition request information to obtain a feature vector of each knowledge graph information associated with each historical knowledge acquisition request information.

[0158] In some embodiments, the agricultural knowledge recommendation model is obtained by training based on the wide and deep interest network model, and the agricultural knowledge recommendation model includes an input layer, a class hidden layer and an output layer; the input layer is used to receive the knowledge acquisition request information and attribute information of the target user; the class hidden layer is used to obtain the target feedback knowledge and the target recommended knowledge category; the output layer is used to output the target feedback knowledge and the target recommended knowledge category

[0159] In some embodiments, the first processing module 510 is also used to receive target input from a target user; based on the target input, determine the target user's knowledge acquisition request information, and obtain the target user's historical usage data on the knowledge platform; based on the target user's registration information, current usage data, and historical usage data on the knowledge platform, determine the target user's attribute information.

[0160] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6 As shown, the electronic device may include: a processor 610, a communication interface 620, a memory 630 and a communication bus 640, wherein the processor 610, the communication interface 620 and the memory 630 communicate with each other through the communication bus 640. The processor 610 may call the logic instructions in the memory 630 to execute the agricultural knowledge recommendation method, which includes: determining the knowledge acquisition request information of the target user and the attribute information of the target user; inputting the knowledge acquisition request information and the attribute information of the target user into the agricultural knowledge recommendation model, and obtaining the target feedback knowledge and the target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model; determining the target recommended knowledge from the agricultural knowledge data set according to the target recommended knowledge category; pushing the target feedback knowledge and the target recommended knowledge to the target user; wherein the agricultural knowledge recommendation model is trained by taking the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and taking the feedback knowledge and the recommended knowledge category corresponding to the historical knowledge acquisition request information as labels.

[0161] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0162] On the other hand, the present invention also provides a computer program product, which includes a computer program, and the computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the agricultural knowledge recommendation method provided by the above methods, and the method includes: determining the knowledge acquisition request information of the target user and the attribute information of the target user; inputting the knowledge acquisition request information and the attribute information of the target user into the agricultural knowledge recommendation model, and obtaining the target feedback knowledge and the target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model; determining the target recommended knowledge from the agricultural knowledge data set according to the target recommended knowledge category; pushing the target feedback knowledge and the target recommended knowledge to the target user; wherein the agricultural knowledge recommendation model is trained using the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and using the feedback knowledge and the recommended knowledge category corresponding to the historical knowledge acquisition request information as labels.

[0163] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the agricultural knowledge recommendation method provided by the above-mentioned methods, the method comprising: determining the knowledge acquisition request information of the target user and the attribute information of the target user; inputting the knowledge acquisition request information and the attribute information of the target user into an agricultural knowledge recommendation model to obtain target feedback knowledge and target recommended knowledge categories associated with the target user output by the agricultural knowledge recommendation model; determining the target recommended knowledge from an agricultural knowledge data set according to the target recommended knowledge categories; pushing the target feedback knowledge and the target recommended knowledge to the target user; wherein the agricultural knowledge recommendation model is trained using the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and using the feedback knowledge and the recommended knowledge categories corresponding to the historical knowledge acquisition request information as labels.

[0164] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0165] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recommending agricultural knowledge, characterized in that: include: Determine knowledge acquisition request information of a target user and attribute information of the target user; Inputting the knowledge acquisition request information and attribute information of the target user into the agricultural knowledge recommendation model to obtain the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model; According to the target recommended knowledge category, determining target recommended knowledge from the agricultural knowledge data set; Pushing the target feedback knowledge and the target recommendation knowledge to the target user; The agricultural knowledge recommendation model is trained by taking the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and taking the feedback knowledge corresponding to the historical knowledge acquisition request information and the recommended knowledge category as labels; the agricultural knowledge data set is determined by the following method: Acquire the attribute information of the user corresponding to each historical knowledge acquisition request information, the feedback knowledge corresponding to each historical knowledge acquisition request information, and the recommended knowledge category; Based on the feedback knowledge corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information is determined; the feedback knowledge corresponding to each historical knowledge acquisition request information is determined based on the knowledge graph information associated with each historical knowledge acquisition request information; Mapping the user's attribute information corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information, and the recommended knowledge category corresponding to each historical knowledge acquisition request information into feature vectors, and combining the feature vectors corresponding to each historical knowledge acquisition request information into a combined feature vector in a preset order; Based on the combined feature vectors corresponding to each historical knowledge acquisition request information, the agricultural knowledge data set is obtained.

2. The agricultural knowledge recommendation method according to claim 1, characterized in that: The user's attribute information includes portrait attribute information and behavior attribute information; the portrait attribute information includes at least one of the user's region, industry, related agricultural crop varieties, gender, age, and education level; the behavior attribute information includes at least one of the time when the user clicked on the knowledge on the knowledge platform, the knowledge category clicked, the reading stay time, whether there is a comment, whether the question is asked by the user, and whether the question is adopted; Mapping the user's attribute information corresponding to each historical knowledge acquisition request information into a feature vector includes: Mapping each attribute information in the portrait attribute information and the behavior attribute information corresponding to each historical knowledge acquisition request information into an attribute sub-vector respectively; All the attribute sub-vectors corresponding to the portrait attribute information and the behavior attribute information corresponding to each historical knowledge acquisition request information are combined to obtain a feature vector of the attribute information of each user corresponding to each historical knowledge acquisition request information.

3. The agricultural knowledge recommendation method according to claim 1, characterized in that: The knowledge graph information associated with the historical knowledge acquisition request information includes at least one of the category of candidate knowledge, knowledge representation form, knowledge source, publishing location, publishing time, number of comments, and whether it has been adopted; Mapping the knowledge graph information associated with each historical knowledge acquisition request information into a feature vector includes: Mapping each type of information in the knowledge graph information associated with each historical knowledge acquisition request information into a knowledge graph sub-vector; All types of knowledge graph sub-vectors in the knowledge graph information associated with each historical knowledge acquisition request information are combined to obtain the feature vectors of each knowledge graph information associated with each historical knowledge acquisition request information.

4. The agricultural knowledge recommendation method according to any one of claims 1 to 3, characterized in that: The agricultural knowledge recommendation model is obtained by training based on a wide and deep interest network model, and the agricultural knowledge recommendation model includes an input layer, a class hidden layer, and an output layer; The input layer is used to receive the knowledge acquisition request information and attribute information of the target user; The class hidden layer is used to obtain the target feedback knowledge and the target recommended knowledge category; The output layer is used to output the target feedback knowledge and the target recommended knowledge category.

5. The agricultural knowledge recommendation method according to any one of claims 1 to 3, characterized in that: The step of determining the knowledge acquisition request information of the target user and the attribute information of the target user includes: receiving a target input from the target user; Based on the target input, determine the knowledge acquisition request information of the target user, and obtain the historical usage data of the target user on the knowledge platform; Based on the registration information, current usage data and historical usage data of the target user on the knowledge platform, attribute information of the target user is determined.

6. An agricultural knowledge recommendation device, characterized in that: include: A first processing module, used to determine knowledge acquisition request information of a target user and attribute information of the target user; The second processing module is used to input the knowledge acquisition request information and attribute information of the target user into the agricultural knowledge recommendation model to obtain the target feedback knowledge and target recommended knowledge category associated with the target user output by the agricultural knowledge recommendation model; A third processing module is used to determine target recommended knowledge from the agricultural knowledge data set according to the target recommended knowledge category; A fourth processing module, used for pushing the target feedback knowledge and the target recommendation knowledge to the target user; The agricultural knowledge recommendation model is trained by taking the historical knowledge acquisition request information of the knowledge platform and the attribute information of the user corresponding to the historical knowledge acquisition request information as samples, and taking the feedback knowledge corresponding to the historical knowledge acquisition request information and the recommended knowledge category as labels; the agricultural knowledge data set is determined by the following method: Acquire the attribute information of the user corresponding to each historical knowledge acquisition request information, the feedback knowledge corresponding to each historical knowledge acquisition request information, and the recommended knowledge category; Based on the feedback knowledge corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information is determined; the feedback knowledge corresponding to each historical knowledge acquisition request information is determined based on the knowledge graph information associated with each historical knowledge acquisition request information; Mapping the user's attribute information corresponding to each historical knowledge acquisition request information, the knowledge graph information associated with each historical knowledge acquisition request information, and the recommended knowledge category corresponding to each historical knowledge acquisition request information into feature vectors, and combining the feature vectors corresponding to each historical knowledge acquisition request information into a combined feature vector in a preset order; Based on the combined feature vectors corresponding to each historical knowledge acquisition request information, the agricultural knowledge data set is obtained.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the agricultural knowledge recommendation method as described in any one of claims 1 to 5 is implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the agricultural knowledge recommendation method as described in any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the agricultural knowledge recommendation method as described in any one of claims 1 to 5 is implemented.

Images