Big data based agricultural information service system

Abstract

The present application relates to the field of agricultural informatization service, especially relates to a kind of agricultural information service system based on big data, the present application is by collecting several user end is applied to the trigger path of target platform, with predetermined period clustering trigger feature corresponding to trigger path;Combining trigger feature and the average trigger number of high-frequency path node, the trigger normality tendency representation value of corresponding trigger path is calculated, to mark trigger path;In response to the marking result, trigger path is analyzed and evaluated;Based on trigger coincidence degree representation value, similar trigger path and trigger path complete time sequence chain are matched;Trigger path is updated clustering, and is arranged in descending order according to trigger number and is shown.The present application can adapt to the non-standard operation habit of agricultural user, ensure that push information and agricultural user actual demand are highly consistent, improve the accuracy of agricultural information service, and simultaneously, improve the operation efficiency of agricultural information platform.

Description

Technical Field

[0001] This invention relates to the field of agricultural information services, and in particular to an agricultural information service system based on big data. Background Technology

[0002] With the deep integration of information technology and the agricultural industry, agricultural informatization has become one of the core driving forces for promoting agricultural modernization. my country's agriculture is upgrading from traditional extensive management to precision and intelligent management. Agricultural production entities, such as farmers, cooperatives, and agricultural enterprises, have an increasingly urgent need for high-quality, personalized agricultural information, covering multiple dimensions such as crop planting techniques and pest and disease control. These needs span the entire agricultural production cycle and exhibit significant characteristics of fragmentation, scenario-based approaches, and high timeliness. Against this backdrop, adapting to the operational characteristics of agricultural users, mining effective demand information from user behavior data, and achieving precise matching between user needs and information services have become key areas for breakthroughs in the current agricultural informatization service field. This also provides a real-world scenario and technological requirement for the development of big data-based agricultural information service systems.

[0003] Chinese Patent Application Publication No. CN119988515A discloses a method and system for agricultural production information navigation based on big data. The system connects a big data platform to the user terminal, extracts data from the platform based on agricultural plot attributes, and generates a global map of the agricultural plot. Based on multi-dimensional real-time monitoring data of the agricultural plot, it determines the operational needs of the agricultural plot, identifies operational information on the global map, and refines the agricultural operations settings for each plot. Based on the user terminal's historical query records on the big data platform, it extracts matching global maps of the agricultural plot, achieving full-link operational information extraction. All extracted global maps of the agricultural plot are integrated and processed to generate full-link agricultural navigation information for the agricultural plot, which is periodically sent to the user terminal, providing users with comprehensive and accurate interactive navigation information for agricultural production.

[0004] However, the following problems still exist in the existing technology.

[0005] Agricultural information is often pushed using broad-spectrum push or simple keyword matching, ignoring the real needs revealed by user behavior, failing to achieve a match between needs, and reducing the accuracy of agricultural information services. Summary of the Invention

[0006] To address this, the present invention provides an agricultural information service system based on big data, which overcomes the problem that existing technologies often use broad-spectrum push or simple keyword matching to push agricultural information, ignoring the real needs presented by user behavior, failing to achieve a matching of needs, and reducing the accuracy of agricultural information services.

[0007] To achieve the above objectives, the present invention provides an agricultural information service system based on big data, comprising:

[0008] The path acquisition module is used to collect several trigger paths of user terminals acting on the target platform, and to cluster the trigger features corresponding to the trigger paths at a predetermined period. The trigger features include the trigger frequency and the number of path nodes involved.

[0009] The trigger analysis module, which is connected to the path acquisition module, is used to combine the trigger characteristics and the average number of triggers of high-frequency path nodes to calculate the trigger normal tendency characterization value of the corresponding trigger path, so as to mark the trigger path.

[0010] A path evaluation module, connected to the trigger analysis module, analyzes and evaluates the trigger path in response to the marking results of the trigger analysis module, including:

[0011] The degree of trigger overlap is evaluated based on the semantic similarity of the keywords corresponding to the trigger path and similar trigger paths, as well as the number of consultation applications corresponding to the trigger path.

[0012] A node analysis module, which is connected to the path evaluation module, is used to match the similar triggering path with the complete temporal chain of the triggering path based on the trigger overlap degree characterization value;

[0013] The update push module, which is connected to the node analysis module, is used to update and cluster the trigger paths and display them in descending order based on the number of triggers.

[0014] Furthermore, the trigger analysis module is used to calculate the trigger normal tendency characterization value of the corresponding trigger path, including:

[0015] The sum of the ratio of trigger frequency to trigger frequency threshold and the ratio of the number of involved path nodes to the number of involved path nodes threshold is used as the first trigger normal feature.

[0016] The ratio of the average number of triggers to the average number of triggers threshold of high-frequency path nodes is used as the second triggering normal characteristic.

[0017] The first triggering normal feature and the second triggering normal feature are weighted and summed to determine the triggering normal tendency characterization value;

[0018] If the trigger frequency of any path node is greater than the trigger frequency threshold, then the path node is determined as the high-frequency path node.

[0019] Furthermore, the trigger analysis module is used to mark the trigger path, including:

[0020] If the triggering normal tendency characterization value of the triggering path is greater than or equal to the triggering normal tendency characterization threshold, then the triggering path is marked.

[0021] Further, the path evaluation module responds to the tagging result of the trigger analysis module, including:

[0022] If any trigger path is marked, then the trigger path is analyzed and evaluated.

[0023] Furthermore, the path evaluation module is used to determine similar triggering paths, including:

[0024] Used to invoke historical trigger paths generated within a predetermined period;

[0025] Used to identify the trigger path and the path nodes corresponding to the historical trigger paths;

[0026] If any historical triggering path meets the path triggering condition, then the historical triggering path is determined as the similar triggering path;

[0027] The path triggering conditions include that the end node is the same as the end node corresponding to the triggering path, and the content browsing time for the end node is not less than the content browsing time threshold.

[0028] Furthermore, the path evaluation module is used to evaluate the trigger overlap degree characterization value, including:

[0029] The ratio of the semantic similarity of the topic words corresponding to the trigger path and similar trigger paths to the semantic similarity threshold is used as the first trigger overlap feature;

[0030] The ratio of the number of consultation requests corresponding to the triggering path to the threshold number of consultation requests is used as the second trigger overlap feature;

[0031] The sum of the first trigger overlap feature and the second trigger overlap feature is used as the trigger overlap degree characterization value.

[0032] Furthermore, the node analysis module is used to match the similar triggering path with the complete temporal chain of the triggering path based on the trigger overlap degree characterization value, including:

[0033] If the trigger overlap characterization value is greater than or equal to the trigger overlap characterization threshold, then the similar trigger paths are matched with the complete timing chain of the trigger paths.

[0034] Furthermore, the node analysis module is used to match the similar triggering path with the complete timing chain of the triggering path, including:

[0035] Based on the temporal position difference between overlapping path nodes and the average topic matching degree of the path nodes corresponding to the path segments between overlapping path nodes, it is determined whether the similar triggering path meets the offset difference fault tolerance benchmark.

[0036] Furthermore, the node analysis module is used to determine whether the similar triggering path meets the offset difference fault tolerance benchmark, including:

[0037] If there exists any similar trigger path and the time sequence difference between the overlapping path nodes corresponding to the trigger path is less than the time sequence difference threshold, and the average topic matching degree is greater than the average topic matching degree threshold, then the similar trigger path is determined to meet the offset difference fault tolerance benchmark.

[0038] Furthermore, the update push module is used to update and cluster the trigger paths, including:

[0039] Cluster similar trigger paths that meet the offset difference fault tolerance benchmark.

[0040] Compared with existing technologies, this invention sets up a path acquisition module to collect trigger paths from several user terminals acting on the target platform, and clusters the trigger features corresponding to the trigger paths at a predetermined period; a trigger analysis module to combine the trigger features and the average number of triggers of high-frequency path nodes to calculate the trigger normal tendency characterization value of the corresponding trigger path, so as to mark the trigger path; a path evaluation module to analyze and evaluate the trigger path in response to the marking results of the trigger analysis module; a node analysis module to match similar trigger paths with the complete temporal chain of trigger paths based on the trigger overlap characterization value; and an update push module to update the clustering of trigger paths and display them in descending order according to the number of triggers.

[0041] In particular, this invention includes a trigger analysis module that analyzes user behavior patterns within a predetermined period to identify effective trigger paths that reflect user needs. This invention comprehensively characterizes the stability and importance of user needs from three dimensions: behavior frequency, path structure, and core nodes. Trigger frequency reflects the degree of user attention to the needs corresponding to the trigger path, quantifying the overall activity and intensity of the trigger path. The number of nodes involved in the corresponding path reflects the complexity of user needs. Furthermore, core nodes within the trigger path are selected, and the dependence of these nodes and the stability of their operational patterns are reflected through extensive user operations. Therefore, this invention calculates a trigger pattern tendency characterization value based on these three characteristics to characterize the normalization of the trigger path and the effectiveness of focusing on core user needs within a predetermined period, providing an efficient and reliable basis for filtering user needs in agricultural information services. Simultaneously, it excludes low-value trigger behaviors such as accidental clicks and misoperations, locking in stable, high-frequency effective demand paths, enabling the system to focus on core user behaviors and improving system operating efficiency.

[0042] In particular, this invention sets up a path evaluation module, which quantifies the overlap between trigger paths through a two-dimensional approach: the relevance of the content of the trigger paths and the urgency of the needs. The semantic similarity of the keywords corresponding to the trigger paths and similar trigger paths reflects the degree of alignment between the trigger paths on the core theme and the relevance of the path content, quantifying the core consistency of the needs and eliminating cases where the paths appear similar but the core needs are unrelated, ensuring the accuracy of the need association. Furthermore, the number of consultation requests corresponding to the trigger paths reflects the urgency of the user's needs, quantifying the actual value of the needs. Users not only browse information but also actively seek further assistance, rather than simply browsing information, making the quantification of the need value more aligned with actual usage scenarios. Therefore, this invention combines the above two features, taking into account both user behavior and need intent, comprehensively reflecting the value of path overlap, and then evaluating the trigger overlap degree representation value. This ensures that the content of the needs is consistent while also demonstrating sufficient demand intensity, jointly quantifying the value of the need overlap between trigger paths. This provides a data foundation for subsequent time-series chain matching, path clustering, and information push, ensuring that the pushed agricultural information not only meets individual user needs but also covers common needs, improving matching accuracy and the targeting and comprehensiveness of services.

[0043] In particular, this invention sets up a node analysis module, which, through a fault-tolerant temporal chain matching mechanism, uncovers deep correlations between triggering paths while ensuring consistency of core needs, providing data support for subsequent clustering and service optimization. It introduces temporal position difference as a measure of fault tolerance, adapting to differences in user operating habits and avoiding omissions of effective matches due to minor differences in the order of path nodes. Simultaneously, it uses topic matching degree as a core constraint to ensure consistency of core user needs. This provides data support for subsequently determining whether similar triggering paths meet the offset difference fault tolerance benchmark. Furthermore, the fault tolerance benchmark enables the system to cope with the diversity and uncertainty of user operations, reducing the impact of abnormal operations on analysis results. It also provides effective samples for the dynamic clustering of the update push module, ensuring that clustering results focus on core needs and avoid interference from irrelevant paths, making the pushed agricultural information align with user operating habits and demand logic, thus improving the targeting of services.

[0044] In particular, this invention includes an update push module that dynamically updates and clusters trigger paths that meet the offset difference tolerance benchmark based on matching results, rather than statically storing them. Furthermore, the information is displayed in descending order of trigger count, enabling users to quickly access popular and high-demand agricultural information, improving user experience, and achieving real-time service iteration. This invention can simultaneously meet the personalized needs of different users, balancing universality and specificity, and ensuring the timeliness and relevance of information pushes. Attached Figure Description

[0045] Figure 1 A functional block diagram of an agricultural information service system based on big data, as an embodiment of the invention;

[0046] Figure 2 This is a logic decision diagram for marking trigger paths in an embodiment of the invention;

[0047] Figure 3 A logic decision diagram for determining similar triggering paths in embodiments of the invention;

[0048] Figure 4 This is a logic diagram for determining whether similar triggering paths meet the offset difference fault tolerance benchmark in an embodiment of the invention. Detailed Implementation

[0049] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0050] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0051] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0052] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral connection; it can refer to a mechanical connection or an electrical connection. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0053] Please see Figure 1 The diagram shown is a functional block diagram of an agricultural information service system based on big data according to an embodiment of the present invention. The agricultural information service system based on big data according to an embodiment of the present invention includes:

[0054] The path acquisition module is used to collect several trigger paths of user terminals acting on the target platform, and to cluster the trigger features corresponding to the trigger paths at a predetermined period. The trigger features include the trigger frequency and the number of path nodes involved.

[0055] The trigger analysis module, which is connected to the path acquisition module, is used to combine the trigger characteristics and the average number of triggers of high-frequency path nodes to calculate the trigger normal tendency characterization value of the corresponding trigger path, so as to mark the trigger path.

[0056] A path evaluation module, connected to the trigger analysis module, analyzes and evaluates the trigger path in response to the marking results of the trigger analysis module, including:

[0057] The degree of trigger overlap is evaluated based on the semantic similarity of the keywords corresponding to the trigger path and similar trigger paths, as well as the number of consultation applications corresponding to the trigger path.

[0058] A node analysis module, which is connected to the path evaluation module, is used to match the similar triggering path with the complete temporal chain of the triggering path based on the trigger overlap degree characterization value;

[0059] The update push module, which is connected to the node analysis module, is used to update and cluster the trigger paths and display them in descending order based on the number of triggers.

[0060] Specifically, the trigger path that the user client can use to act on the target platform can be obtained from the server logs of the target platform with the user client's permission, which will not be elaborated further.

[0061] It is understandable that, in order to ensure the validity of the collected data and to align with the operational habits of agricultural users, such as the fact that agricultural production activities, including sowing, fertilization, and pest and disease control, are often planned on a weekly basis, and that agricultural information services, such as pest and disease warnings and agricultural technology guidance programs, are updated weekly, this embodiment sets the predetermined period to 7 days to achieve a balance between data validity, timeliness of demand, and accuracy of service. Further details will not be elaborated upon here.

[0062] Specifically, the complete temporal chain refers to the complete behavioral trajectory sequence generated from "initial operation triggering" to "end node completion" when the user terminal acts on the agricultural information service target platform, which includes all continuous path nodes and the temporal sequence relationship between path nodes.

[0063] Specifically, there are no restrictions on the specific structure of the path acquisition module, trigger analysis module, path evaluation module, node analysis module, and update push module. Each module or its units can be composed of logical components or combinations of logical components. Logical components include field-programmable processors, computers, or microprocessors in computers.

[0064] Specifically, the trigger analysis module is used to calculate the trigger normality tendency characterization value of the corresponding trigger path, including:

[0065] The sum of the ratio of trigger frequency to trigger frequency threshold and the ratio of the number of involved path nodes to the number of involved path nodes threshold is used as the first trigger normal feature.

[0066] The ratio of the average number of triggers to the average number of triggers threshold of high-frequency path nodes is used as the second triggering normal characteristic.

[0067] The first triggering normal feature and the second triggering normal feature are weighted and summed to determine the triggering normal tendency characterization value;

[0068] If the trigger frequency of any path node is greater than the trigger frequency threshold, then the path node is determined as the high-frequency path node.

[0069] Specifically, the purpose of weighting the first and second triggering normality features is to accurately identify high-value paths that are stable, frequent, and focused on core user needs, rather than simply active or complex paths. The average number of triggers for high-frequency path nodes more directly reflects the stability and focus of core user needs, eliminating interference from non-core nodes and accurately targeting core user requirements. For example, key steps in agricultural information queries such as "disease diagnosis" and "fertilization plans" are core criteria for determining the normality of a path. The two features involved in the first triggering normality feature primarily reflect the activity level and operational complexity of the triggering path. For instance, user errors might lead to a high trigger frequency and many path nodes for a particular path, but fewer triggers for core nodes. In this case, the first triggering normality feature value is high, but the actual value of the path is low. Furthermore, the core of the agricultural information service system is matching the actual needs of users; the operational stability of core nodes corresponding to the second normality triggering feature, compared to the overall activity level of the triggering path, better reflects the authenticity and normality of user needs. Therefore, the second triggering normality feature, calculated based on the average number of triggers for high-frequency path nodes, is assigned a higher weight coefficient to identify high-value paths and provide a more reliable basis for subsequent service optimization; this coefficient is set to 0.6. Correspondingly, the weight coefficient of the first triggering normality feature, calculated based on triggering features, i.e., trigger frequency and the number of path nodes involved, is set to 0.4.

[0070] In this embodiment, setting the trigger frequency threshold and the threshold for the number of path nodes involved are both intended to characterize situations where the complexity of user-side requirements is high. The average trigger count threshold is intended to characterize situations where the stability of core user-side requirements is high and the focus of requirements is clear. By acquiring relevant data from several historical predetermined periods, and calling historical trigger frequency data of trigger paths, historical data of the number of path nodes involved, and historical data of the average trigger count of high-frequency path nodes, the average trigger frequency, the average number of path nodes involved, and the average trigger count are calculated and used as the baseline values ​​under normal circumstances. The purpose of the three thresholds is to determine the trigger frequency threshold as the product of the average trigger frequency and the first deviation coefficient, the number of involved path nodes threshold as the product of the average number of involved path nodes and the second deviation coefficient, and the average number of triggers threshold as the product of the average number of triggers and the third deviation coefficient. The first deviation coefficient is selected within the interval [1.2, 1.4], preferably 1.2 in practice; the second deviation coefficient is selected within the interval [1.2, 1.3], preferably 1.3 in practice; and the third deviation coefficient is selected within the interval [1.4, 1.6], preferably 1.4 in practice.

[0071] The purpose of setting the trigger frequency threshold is to characterize the situation where the path node can meet the needs of most users. By obtaining relevant data from several historical predetermined periods, calling the historical trigger frequency data of the path node, the average trigger frequency is calculated and used as a benchmark value under normal circumstances. Based on the purpose of setting the trigger frequency threshold, the trigger frequency threshold is determined as the product of the average trigger frequency and the frequency deviation coefficient. The frequency deviation coefficient is selected in the interval [1.4, 1.6], and is preferably 1.4 in practice.

[0072] Specifically, this invention includes a trigger analysis module that analyzes user behavior patterns within a predetermined period to identify effective trigger paths that reflect user needs. This invention comprehensively characterizes the stability and importance of user needs from three dimensions: behavior frequency, path structure, and core nodes. Trigger frequency reflects the degree of user attention to the needs corresponding to the trigger path; a higher frequency indicates a stronger and more stable need behind the path, quantifying the overall activity and intensity of the trigger path. The number of corresponding path nodes reflects the complexity of user needs; more path nodes generally mean more complex user needs, while fewer path nodes indicate more direct and simple user needs. Furthermore, core nodes within the trigger path are selected, and the dependence and operational stability of these nodes are reflected through extensive user operations; a higher average number of triggers indicates that the high-frequency trigger node can focus on the needs of a large number of users. Therefore, this invention calculates a trigger normalization tendency characterization value based on these three characteristics to characterize the normalization of the trigger path and the effectiveness of focusing on core user needs within a predetermined period, providing an efficient and reliable basis for user need screening in agricultural information services. At the same time, by eliminating low-value triggering behaviors such as accidental clicks and misoperations, the system focuses on stable, high-frequency, and effective demand paths, enabling it to concentrate on core user behaviors and improve system operating efficiency.

[0073] Specifically, please refer to Figure 2 As shown, this is a logic decision diagram for marking trigger paths according to an embodiment of the present invention. The trigger analysis module is used to mark the trigger paths, including:

[0074] If the triggering normal tendency characterization value of the triggering path is greater than or equal to the triggering normal tendency characterization threshold, then the triggering path is marked.

[0075] If the triggering normal tendency characterization value of the triggering path is less than the triggering normal tendency characterization threshold, then there is no need to mark the triggering path.

[0076] The triggering normal tendency characterization threshold is predetermined. The triggering normal tendency characterization value is determined by calculating the triggering frequency equal to the triggering frequency threshold, the number of involved path nodes equal to the number of involved path nodes threshold, and the average number of triggers of high-frequency path nodes equal to the average number of triggers threshold.

[0077] Specifically, the path evaluation module responds to the tagging result of the trigger analysis module, including:

[0078] If any trigger path is marked, then the trigger path is analyzed and evaluated.

[0079] Specifically, please refer to Figure 3 As shown, this is a logic decision diagram for determining similar triggering paths in an embodiment of the present invention. The path evaluation module is used to determine similar triggering paths and includes:

[0080] Used to invoke historical trigger paths generated within a predetermined period;

[0081] Used to identify the trigger path and the path nodes corresponding to the historical trigger paths;

[0082] If any historical triggering path meets the path triggering condition, then the historical triggering path is determined as the similar triggering path;

[0083] The path triggering conditions include that the end node is the same as the end node corresponding to the triggering path, and the content browsing time for the end node is not less than the content browsing time threshold.

[0084] In this embodiment, the purpose of setting a content browsing duration threshold is to characterize the situation where the user's attention to the content corresponding to the terminal node is high and the degree of matching between their needs and the content is high. By obtaining relevant data from several historical predetermined periods, calling the historical data of content browsing duration of the terminal node, the average content browsing duration is calculated and used as a benchmark value under normal circumstances. Based on the purpose of setting the content browsing duration threshold, the content browsing duration threshold is determined as the product of the average content browsing duration and the duration deviation coefficient. The duration deviation coefficient is selected in the interval [1.2, 1.3], and is preferably 1.2 in practice.

[0085] Specifically, the path evaluation module is used to evaluate the trigger overlap degree characterization value, including:

[0086] The ratio of the semantic similarity of the topic words corresponding to the trigger path and similar trigger paths to the semantic similarity threshold is used as the first trigger overlap feature;

[0087] The ratio of the number of consultation requests corresponding to the triggering path to the threshold number of consultation requests is used as the second trigger overlap feature;

[0088] The sum of the first trigger overlap feature and the second trigger overlap feature is used as the trigger overlap degree characterization value.

[0089] In this embodiment, the purpose of setting a threshold for the number of consultation requests is to characterize situations where the user's needs are more urgent. By obtaining relevant data from several historical predetermined periods, calling the historical data of the number of consultation requests corresponding to the trigger path, the average number of consultation requests is calculated and used as a benchmark value under normal circumstances. Based on the purpose of setting the threshold for the number of consultation requests, the threshold for the number of consultation requests is determined as the product of the average number of consultation requests and the number deviation coefficient. The number deviation coefficient is selected within the interval [1.3, 1.4], and is preferably 1.3 in practice.

[0090] Specifically, the purpose of setting the semantic similarity threshold is to characterize the high degree of semantic association between the core topic words corresponding to the value triggering path and the similar triggering path, reflecting the high degree of fit between user needs. Based on this, the semantic similarity threshold is set to 0.7.

[0091] The semantic similarity between trigger paths and corresponding topic words of similar trigger paths is determined by calculating cosine similarity. Correspondingly, the average topic matching degree can also be determined by calculating cosine similarity, which will not be elaborated here.

[0092] Specifically, this invention includes a path evaluation module that quantifies overlap in two dimensions: the content relevance of the trigger path and the urgency of the need. The semantic similarity between the trigger path and similar trigger paths reflects the degree of alignment between the trigger paths on their core themes and the relevance of their content, quantifying the core consistency of the need. Higher semantic similarity indicates a more consistent core content among user needs, such as both revolving around the same agricultural theme like "wheat pest and disease control" or "vegetable greenhouse temperature control technology." This eliminates cases where paths appear similar but the core needs are unrelated, ensuring the accuracy of need association. Furthermore, the number of consultation requests corresponding to each trigger path reflects the urgency of the user's need, quantifying the actual value of the need. More consultation requests indicate a more urgent and practically valuable need, suggesting that users are not merely browsing information but actively seeking further assistance, rather than simply reading information. This makes the quantification of need value more aligned with actual usage scenarios. Therefore, this invention combines the above two features, taking into account both user behavior and intent, to comprehensively reflect the value of path overlap, and then evaluate the degree of trigger overlap. This ensures consistency in demand content while also demonstrating sufficient demand intensity, collectively quantifying the value of demand overlap between trigger paths. This provides a data foundation for subsequent time-series chain matching, path clustering, and information push, ensuring that the pushed agricultural information not only meets individual user needs but also covers common needs, improving matching accuracy and the targeted and comprehensive nature of the service.

[0093] Specifically, the node analysis module is used to match the similar triggering path with the complete temporal chain of the triggering path based on the trigger overlap degree characterization value, including:

[0094] If the trigger overlap degree characterization value is greater than or equal to the trigger overlap degree characterization threshold, then the similar trigger path is matched with the complete timing chain of the trigger path.

[0095] The trigger overlap degree characterization threshold is predetermined. The trigger overlap degree characterization value calculated is determined when the semantic similarity of the topic words corresponding to the trigger path and similar trigger paths is equal to the semantic similarity threshold, and the number of consultation applications corresponding to the trigger path is equal to the number of consultation applications threshold.

[0096] Specifically, the node analysis module is used to match the similar triggering path with the complete timing chain of the triggering path, including:

[0097] Based on the temporal position difference between overlapping path nodes and the average topic matching degree of the path nodes corresponding to the path segments between overlapping path nodes, it is determined whether the similar triggering path meets the offset difference fault tolerance benchmark.

[0098] Specifically, please refer to Figure 4 As shown, this is a logic diagram for determining whether a similar triggering path meets the offset difference fault tolerance benchmark in an embodiment of the present invention. The node analysis module is used to determine whether the similar triggering path meets the offset difference fault tolerance benchmark, including:

[0099] If there exists any similar trigger path and the time sequence difference between the overlapping path nodes corresponding to the trigger path is less than the time sequence difference threshold, and the average topic matching degree is greater than the average topic matching degree threshold, then the similar trigger path is determined to meet the offset difference fault tolerance benchmark.

[0100] In this embodiment, the purpose of setting the temporal position difference threshold and the average topic matching degree threshold is to characterize the situation where the core needs of the user end corresponding to the trigger path and similar trigger paths are relatively close. By obtaining relevant data of several historical predetermined periods, calling historical data of temporal position difference and historical data of average topic matching degree, the mean of temporal position difference and the mean of average topic matching degree are calculated, and the corresponding values ​​are used as the benchmark values ​​under normal circumstances. Based on the purpose of setting the above two thresholds, the temporal position difference threshold is determined as the product of the mean of temporal position difference and the first offset coefficient, and the average topic matching degree threshold is determined as the product of the mean of average topic matching degree and the second offset coefficient. The first offset coefficient is selected in the interval [0.9, 0.95], preferably 0.9 in the implementation, and the second offset coefficient is selected in the interval [1.15, 1.25], preferably 1.15 in the implementation.

[0101] Specifically, the average topic matching is used to measure the semantic fit of topic words of each path node in the continuous path segment defined by overlapping path nodes in the similar triggering path and the triggering path in chronological order.

[0102] Specifically, the absolute value of the maximum difference between the sequence number of the overlapping path node in the timing chain corresponding to the similar trigger path and the sequence number of the overlapping path node in the timing chain corresponding to the trigger path is taken as the timing position difference, which will not be elaborated further.

[0103] Specifically, this invention establishes a node analysis module that, through a fault-tolerant temporal chain matching mechanism, uncovers deep correlations between triggering paths while ensuring consistency in core requirements, providing data support for subsequent clustering and service optimization. It introduces temporal position difference as a fault-tolerant benchmark to adapt to differences in user operating habits, such as adjusting the order of some steps, avoiding omissions of effective matching due to minor differences in the order of path nodes. Simultaneously, it uses topic matching degree as a core constraint to ensure consistency in core user needs. This provides data support for subsequently determining whether similar triggering paths meet the offset difference fault-tolerant benchmark. Furthermore, the fault-tolerant benchmark enables the system to cope with the diversity and uncertainty of user operations, such as path correction after accidental clicks or repeated steps, reducing the impact of abnormal operations on analysis results. Moreover, it provides effective samples for the dynamic clustering of the update push module, ensuring that clustering results focus on core needs and avoid interference from irrelevant paths, making the pushed agricultural information align with user operating habits and demand logic, thus improving the targeting of services.

[0104] Specifically, the update push module is used to update and cluster the trigger paths, including:

[0105] Cluster similar trigger paths that meet the offset difference fault tolerance benchmark.

[0106] Specifically, this invention includes an update push module that dynamically updates and clusters trigger paths that meet the offset difference tolerance benchmark based on matching results, rather than statically storing them. Furthermore, it displays path nodes in descending order of trigger count, enabling users to quickly access popular and high-demand agricultural information, improving user experience, and achieving real-time service iteration. This invention can simultaneously meet the personalized needs of different users, balancing universality and specificity, and ensuring the timeliness and relevance of information pushes.

[0107] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An agricultural information service system based on big data, characterized in that, include: The path acquisition module is used to collect several trigger paths of user terminals acting on the target platform, and to cluster the trigger features corresponding to the trigger paths at a predetermined period. The trigger features include the trigger frequency and the number of path nodes involved. The trigger analysis module is used to combine the trigger characteristics and the average number of triggers of high-frequency path nodes to calculate the trigger normal tendency characterization value of the corresponding trigger path, so as to mark the trigger path. A path evaluation module, in response to the marking results of the trigger analysis module, analyzes and evaluates the trigger path, including: The degree of trigger overlap is evaluated based on the semantic similarity of the keywords corresponding to the trigger path and similar trigger paths, as well as the number of consultation applications corresponding to the trigger path. The node analysis module is used to match the similar triggering path with the complete temporal chain of the triggering path based on the trigger overlap degree characterization value; The update push module is used to update and cluster the trigger paths, and then display them in descending order based on the number of triggers. The trigger analysis module is used to calculate the trigger normal tendency characterization value of the corresponding trigger path, including: The sum of the ratio of trigger frequency to trigger frequency threshold and the ratio of the number of involved path nodes to the number of involved path nodes threshold is used as the first trigger normal feature. The ratio of the average number of triggers to the average number of triggers threshold of high-frequency path nodes is used as the second triggering normal characteristic. The first triggering normal feature and the second triggering normal feature are weighted and summed to determine the triggering normal tendency characterization value; If the trigger frequency of any path node is greater than the trigger frequency threshold, then the path node is determined as the high-frequency path node. The path evaluation module is used to evaluate the trigger overlap degree characterization value, including: The ratio of the semantic similarity of the topic words corresponding to the trigger path and similar trigger paths to the semantic similarity threshold is used as the first trigger overlap feature; The ratio of the number of consultation requests corresponding to the triggering path to the threshold number of consultation requests is used as the second trigger overlap feature; The sum of the first trigger overlap feature and the second trigger overlap feature is used as the trigger overlap degree characterization value; The node analysis module is used to determine whether the similar triggering path meets the offset difference fault tolerance benchmark, including: If there exists any similar trigger path and the time sequence difference between the overlapping path nodes corresponding to the trigger path is less than the time sequence difference threshold, and the average topic matching degree is greater than the average topic matching degree threshold, then the similar trigger path is determined to meet the offset difference fault tolerance benchmark.

2. The agricultural information service system based on big data according to claim 1, characterized in that, The trigger analysis module is used to mark the trigger path, including: If the triggering normal tendency characterization value of the triggering path is greater than or equal to the triggering normal tendency characterization threshold, then the triggering path is marked.

3. The agricultural information service system based on big data according to claim 2, characterized in that, The path evaluation module responds to the tagging result of the trigger analysis module, including: If any trigger path is marked, then the trigger path is analyzed and evaluated.

4. The agricultural information service system based on big data according to claim 1, characterized in that, The path evaluation module is used to determine similar triggering paths, including: Used to invoke historical trigger paths generated within a predetermined period; Used to identify the trigger path and the path nodes corresponding to the historical trigger paths; If any historical triggering path meets the path triggering condition, then the historical triggering path is determined as the similar triggering path; The path triggering conditions include that the end node is the same as the end node corresponding to the triggering path, and the content browsing time for the end node is not less than the content browsing time threshold.

5. The agricultural information service system based on big data according to claim 1, characterized in that, The node analysis module is used to match the similar triggering paths with the complete temporal chain of the triggering paths based on the trigger overlap degree characterization value, including: If the trigger overlap degree characterization value is greater than or equal to the trigger overlap degree characterization threshold, then the similar trigger path is matched with the complete timing chain of the trigger path.

6. The agricultural information service system based on big data according to claim 1, characterized in that, The node analysis module is used to match the similar trigger paths with the complete timing chain of the trigger paths, including: Based on the temporal position difference between overlapping path nodes and the average topic matching degree of the path nodes corresponding to the path segments between overlapping path nodes, it is determined whether the similar triggering path meets the offset difference fault tolerance benchmark.

7. The agricultural information service system based on big data according to claim 1, characterized in that, The update push module is used to update and cluster the trigger paths, including: Cluster similar trigger paths that meet the offset difference fault tolerance benchmark.

Images