A method for analyzing the hot spots of policy and academic research in the Yellow River Basin

Abstract

The application provides a hot dynamic analysis method for policy and academic research of the Yellow River Basin, comprising the following steps: step 1, retrieving mining samples related to the policy and academic research of the Yellow River Basin from a preset database; step 2, pre-analyzing the mining samples based on a preset feature set to determine the network structure of the information entity of the policy and academic system of the Yellow River Basin; and step 3, performing feature analysis on the network structure to obtain a hot dynamic analysis result and outputting the result. The hot dynamic of different basins is determined by integrating and analyzing the academic and policy characteristics from multiple dimensions, and the efficiency of evolution research is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of basin hotspot analysis, and particularly relates to a hotspot dynamic analysis method for policy and academic research on the Yellow River basin. BACKGROUND

[0002] The ecological protection and high-quality development of the Yellow River basin has become a national strategy, and the research hotspots and frontiers in the basin are increasingly concerned. The Yellow River problem is relatively complex, and the natural conditions of different regions in the upper, middle and lower reaches of the Yellow River are different, and the protection and management priorities of the basin are different. Clarifying the historical research hotspots and accurately judging the frontiers of research are important supports for the systematic management and overall management of the upper, middle and lower reaches of the Yellow River, but the current research results mainly take "Yellow River" as the academic literature retrieval theme, and use citespace and other software to macroscopically analyze the research situation of the Yellow River basin high-quality development in the academic field. These results lack systematic analysis of regional key hot issues and management measures.

[0003] In order to improve the pertinence and effectiveness of policies and engineering measures in the upper, middle and lower reaches of the Yellow River, it is necessary to classify and analyze the characteristics of the upper, middle and lower reaches of the Yellow River, and then effectively promote the overall protection and management of the Yellow River basin. The existing research analyzes the problems and status of the upper, middle and lower reaches of the Yellow River, but needs to pay attention to the time as a universal and important dimension to measure the development trend of theory, and at the same time, the change of national policy is also one of the important factors to promote the change of academic frontiers of social science, that is, the mutual feedback and interaction between academic frontiers and national policy is an important reflection of the development trend of the Yellow River basin. The evolution of the academic situation and policy status of the upper, middle and lower reaches of the Yellow River is systematically lacking in a long period of time.

[0004] Therefore, the present application provides a hotspot dynamic analysis method for policy and academic research on the Yellow River basin. SUMMARY

[0005] The present application provides a hotspot dynamic analysis method for policy and academic research on the Yellow River basin, which integrates the analysis of academic and policy from multiple dimension characteristics to determine the hotspot dynamics of different basins, and improves the efficiency of evolution research.

[0006] The present application provides a hotspot dynamic analysis method for policy and academic research on the Yellow River basin, which integrates the analysis of academic and policy from multiple dimension characteristics to determine the hotspot dynamics of different basins, and improves the efficiency of evolution research.

[0007] Step 1: retrieving mining samples related to policy and academic research on the Yellow River basin from a preset database;

[0008] Step 2: based on a preset feature set, pre-analyzing the mining samples to determine the network structure of the policy and academic system information entity of the Yellow River basin;

[0009] Step 3: feature analysis is performed on the network structure, and a hotspot dynamic analysis result is obtained and output.

[0010] In a possible implementation, the preset feature set includes: a text node degree feature, an intermediary centrality feature, a text information entropy feature, and a text burstiness feature.

[0011] In a possible implementation, step 2: based on the preset feature set, the mining sample is pre-analyzed to determine the network structure of the policy and academic system information entity in the Yellow River Basin, including:

[0012] The text node degree of the academic articles and the policy articles in the mining sample is determined, and then the influence of different nodes is determined;

[0013] The intermediary centrality of the network nodes corresponding to the academic articles and the policy articles in the mining sample is determined, and based on the intermediary centrality, the application scenarios of different articles and the control ability of the articles on the keyword network are determined;

[0014] The text information entropy of the academic articles and the policy articles in the mining sample is determined, and then the associated information of high-frequency keywords is determined;

[0015] The text burstiness feature of the academic articles and the policy articles in the mining sample is determined, and then the research frontiers and hotspots of the Yellow River Basin are determined;

[0016] Based on all the determination results, the network structure of the policy and academic system information entity in the Yellow River Basin is obtained;

[0017] The mining sample includes: academic articles and policy articles.

[0018] In a possible implementation, determining the text burstiness feature of the mining sample includes:

[0019] The sample publication time domain of the academic articles in the mining sample is obtained, and based on the sample publication time domain, the policy article acquisition time domain is determined;

[0020] The policy articles in the acquisition time domain are obtained;

[0021] All sample keywords of the academic articles and all policy keywords of the policy articles in the Yellow River Basin are determined;

[0022] Based on the time stamp, the sample keywords and the policy keywords corresponding to the same time are burstiness processed, and the burstiness time lengths of the associated keywords in the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River are sorted based on the burstiness processing result;

[0023] Meanwhile, based on the ranking result, the continuity of the burst phenomenon of different associated keywords is calibrated, and then the text burst characteristics are determined.

[0024] In a possible implementation, after the policy articles between the acquisition time domains are acquired, the following steps are included:

[0025] The academic articles and the policy articles are subjected to time series analysis to determine the article research trends of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River.

[0026] The article publishing agencies of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River are recorded, and then the article publishing river distribution of each agency, the author publishing distribution of each author, and the publishing cited distribution are acquired.

[0027] Based on the article publishing river distribution, the author publishing distribution, the publishing cited distribution, and the article research trend, the research fields and research objects of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River are determined.

[0028] The first N1 cited keywords related to the academic articles in the publishing cited distribution are acquired and analyzed, and the network node number, the edge number, and the network density of the academic literature keywords of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River are acquired, and then the first cited cluster of the first cited keyword of each basin is determined.

[0029] The first N2 cited keywords related to the policy articles in the publishing cited distribution are acquired and analyzed, and the network node number, the edge number, and the network density of the policy keywords of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River are acquired, and then the second cited cluster of the second cited keyword of each basin is determined.

[0030] Based on the first cited cluster and the second cited cluster, the keyword co-occurrence result of the Yellow River basin is determined.

[0031] Based on the research fields, the research objects, and the keyword co-occurrence result, the corresponding research hotspots are determined.

[0032] In a possible implementation, after the text node degree of the academic articles and the policy articles in the mining sample is determined, the following steps are included:

[0033] The text structure constituted by the mining sample is determined, and the degree structure is constructed according to the acquired text node degree.

[0034] The text structure and the degree structure are subjected to consistency analysis, when the consistency analysis result meets the preset division condition, the mining sample is subjected to primary division according to the text node degree, and the corresponding division mode is called from the node-division database.

[0035] determine the compactness between different division blocks after initial division, determine the division accuracy of the division result based on the compactness, when the division accuracy is lower than a preset accuracy, obtain a first participation parameter for determining the compactness, and simultaneously, obtain a second participation parameter in the division process and a division parameter of the division mode;

[0036] based on the first participation parameter, the second participation parameter and the division parameter, construct a parameter matrix, and according to a parameter attribute unification principle, first calibrate consistent parameters in the parameter matrix, and second calibrate inconsistent parameters;

[0037] Meanwhile, determine the difference parameter of the first calibration parameter and the second calibration parameter based on the standard parameter of the preset accuracy, adjust the division mode according to the difference parameter, and perform secondary division on the mining sample according to the adjusted mode, so as to realize the construction of the network structure.

[0038] In a possible implementation manner, after determining the corresponding research hotspots, the method further comprises:

[0039] monitor the location of the Yellow River Basin, and obtain a difficult monitoring map and a simple monitoring map according to the location monitoring result;

[0040] obtain a first monitoring sample corresponding to the difficult monitoring map in history, and determine a first monitoring location point according to a location classification model, and simultaneously, perform map analysis on the difficult monitoring map to determine the must monitoring location points that exist, if the must monitoring location points are missing in the first monitoring location points, obtain the point location information of the missing monitoring points, and determine the survey difficulty according to the point location information, and match a subsequent survey mode related to the survey difficulty;

[0041] continuously survey the missing monitoring points according to the subsequent survey mode, and take the missing monitoring points as the monitoring sample corresponding to the difficult monitoring map, and perform first analysis on all the samples after continuous surveying according to a monitoring index classification model, to determine the same type of monitoring location points corresponding to the same type of monitoring index in the difficult monitoring map;

[0042] obtain a second monitoring sample corresponding to the simple monitoring map in history, and perform second analysis on the second monitoring sample according to a monitoring index classification model, to determine the same type of monitoring location points corresponding to the same type of monitoring index in the simple monitoring map;

[0043] determine the location distribution of the same type of monitoring location points, and set a first label, and the first label comprises: a location point and an index allocation related to the location point;

[0044] Split the research hotspots to obtain a first hot spot set in the upper reaches of the Yellow River, set a first hot spot label, obtain a second hot spot set in the middle reaches of the Yellow River, set a second hot spot label, obtain a third hot spot set in the lower reaches of the Yellow River, and set a third hot spot label;

[0045] Construct a label relationship between the first label, the first hot spot label, the second hot spot label, and the third hot spot label;

[0046] When the label relationship meets a preset hot spot distribution relationship, the corresponding research hot spot is kept unchanged;

[0047] When the label relationship does not meet the preset hot spot distribution relationship, the research hot spot is expanded based on the label information of the first label, and the obtained research hot spot is taken as a common research hot spot.

[0048] In a possible implementation manner, in step 3, the network structure is subjected to feature analysis, a hot spot dynamic analysis result is obtained, and output is performed, including:

[0049] Obtain a structure parameter involved in the network structure;

[0050] Based on a feature analysis model, the structure parameter is analyzed to obtain a feature sequence, and a sequence observation point between the feature sequence is determined;

[0051] According to the sequence observation point of the same sequence, the sequence bias of the feature sequence is determined, and the index bias of a feature index related to the feature analysis model is uniformly processed to determine whether a uniform processing result meets a preset uniform condition. If yes, the feature sequence corresponding hot spot dynamics are determined based on the feature index, and output is performed;

[0052] Otherwise, according to the basin information of different basins, the sequence observation points of different sequences are classified into the same basin, and according to the classified result, hot spot dynamics of the same basin are obtained, and output is performed.

[0053] In a possible implementation manner, the hot spot dynamic analysis result is obtained, and output is performed, further including:

[0054] Obtain a hot spot dynamic analysis atlas of different basins in the Yellow River basin;

[0055] Obtain a sudden impact event corresponding to different basins in the Yellow River basin, and determine an actual impact index of the sudden impact event on the hot spot dynamics based on the event impact model;

[0056] Determine an actual impact value Y1 according to the actual impact index;

[0057]

[0058] wherein, S represents the number of the actual influence indicators; d s1 represents the index value of the s1th actual influence indicator after standard numerical conversion, and the value range is [0, 1]; f s1 (D1, d s1 ) represents the influence coefficient of the s1th actual influence indicator on the hotspot dynamic distribution map D1 corresponding to the upper reaches of the Yellow River; h1 represents the position coefficient of the upper reaches of the Yellow River and the burst location of the burst influence event; h2 represents the position coefficient of the middle reaches of the Yellow River and the burst location of the burst influence event; h3 represents the position coefficient of the lower reaches of the Yellow River and the burst location of the burst influence event; f s1 (D2, d s1 ) represents the influence coefficient of the s1th actual influence indicator on the hotspot dynamic distribution map D2 corresponding to the middle reaches of the Yellow River; f s1 (D3, d s1 ) represents the influence coefficient of the s1th actual influence indicator on the hotspot dynamic distribution map D3 corresponding to the lower reaches of the Yellow River;

[0059] When the actual influence value Y1 is greater than the preset value, the maps with influence coefficients greater than the preset coefficient are selected from f s1 (D1, d s1 ), f s1 (D2, d s1 ), and f s1 (D3, d s1 ), and the corresponding selected maps are adjusted based on the burst influence event, to obtain new maps, which are output;

[0060] When the actual influence value Y1 is less than the preset value, the corresponding actual influence indicator is regarded as an invalid indicator, and the original output map content remains unchanged.

[0061] In a possible implementation manner, in the process of step 1: retrieving the mining samples related to the policy and academic research of the Yellow River basin from the preset database, the process includes:

[0062] retrieving a first sample related to the Yellow River basin theme from a preset database;

[0063] performing basin correlation analysis on each academic article in the first sample, retaining the academic articles with a correlation degree greater than a first preset degree, deleting the academic articles with a correlation degree less than a second preset degree, and retaining the academic articles with a correlation degree between the first preset degree and the second preset degree;

[0064] extracting academic keywords in each academic article to be retained, determining a first number of the academic keywords, determining research keywords in the academic keywords, and determining a second number of the research keywords;

[0065] establishing a first academic graph of the research keywords and a second academic graph of the academic keywords, and determining academic overlap of the first academic graph based on the second academic graph;

[0066]

[0067] wherein F1 represents the academic overlap; n1 represents the number of research keywords; n2 represents the number of remaining keywords in the academic keywords after removing the research keywords; n3 represents the knowledge line related to each research keyword; n4 represents the knowledge line related to each remaining keyword; A2 i1,j1 represents the knowledge proportion of the j1th knowledge line related to the i1th research keyword based on the corresponding academic article; A1 i1 represents the knowledge proportion of the i1th research keyword based on the corresponding academic article; A3 i2 represents the knowledge proportion of the i2th remaining keyword based on the corresponding academic article; A4 i2,j2 represents the knowledge proportion of the j2th knowledge line related to the i2th remaining keyword based on the corresponding academic article;

[0068] when the second number and the number ratio of the second number are greater than a preset ratio or the academic overlap is greater than a preset overlap, performing a retention operation on the corresponding academic article to be retained;

[0069] based on the retained academic article sample and the policy article obtained in the corresponding time domain to constitute a mining sample.

[0070] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be achieved and obtained by the structure particularly pointed out in the written description and the accompanying drawings.

[0071] The technical solutions of the present application will be further described in detail below with the help of the accompanying drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0072] The accompanying drawings are intended to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0073] Figure 1 is a flow chart of a policy and academic research hotspot dynamic analysis method in the embodiment of the present application in the Yellow River Basin;

[0074] Figure 2-1 is an academic author co-occurrence and clustering graph in the upper reaches of the Yellow River in the embodiment of the present application;

[0075] Figure 2-2 The co-occurrence and clustering graph of academic authors in the middle reaches of the Yellow River in the embodiment of the present application;

[0076] Figure 2-3 The co-occurrence and clustering graph of academic authors in the lower reaches of the Yellow River in the embodiment of the present application;

[0077] Figure 3-1 The co-occurrence and clustering graph of academic keywords in the upper reaches of the Yellow River in the embodiment of the present application;

[0078] Figure 3-2 The co-occurrence and clustering graph of academic keywords in the middle reaches of the Yellow River in the embodiment of the present application;

[0079] Figure 3-3 The co-occurrence and clustering graph of academic keywords in the lower reaches of the Yellow River in the embodiment of the present application;

[0080] Figure 4-1 The co-occurrence and clustering graph of policy keywords in the upper reaches of the Yellow River in the embodiment of the present application;

[0081] Figure 4-2 The co-occurrence and clustering graph of policy keywords in the middle reaches of the Yellow River in the embodiment of the present application;

[0082] Figure 4-3 The co-occurrence and clustering graph of policy keywords in the lower reaches of the Yellow River in the embodiment of the present application;

[0083] Figure 5-1 The dynamic analysis graph of academic research hotspots in the upper reaches of the Yellow River in the embodiment of the present application;

[0084] Figure 5-2 The dynamic analysis graph of policy research hotspots in the upper reaches of the Yellow River in the embodiment of the present application;

[0085] Figure 6-1 The dynamic analysis graph of academic research hotspots in the middle reaches of the Yellow River in the embodiment of the present application;

[0086] Figure 6-2 The dynamic analysis graph of policy research hotspots in the middle reaches of the Yellow River in the embodiment of the present application;

[0087] Figure 7-1 The dynamic analysis graph of academic research hotspots in the lower reaches of the Yellow River in the embodiment of the present application;

[0088] Figure 7-2 The dynamic analysis graph of policy research hotspots in the lower reaches of the Yellow River in the embodiment of the present application. DETAILED DESCRIPTION

[0089] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to explain and illustrate the present application, and are not used to limit the present application.

[0090] Embodiment 1

[0091] The application provides a hotspot dynamic analysis method for policy and academic research in the Yellow River Basin, as shown in the following formula (I) : Figure 1 The method comprises the following steps of:

[0092] Step 1: retrieving mining samples related to policy and academic research in the Yellow River Basin from a preset database;

[0093] Step 2: pre-analyzing the mining samples based on a preset feature set to determine the network structure of information entities of the policy and academic system in the Yellow River Basin;

[0094] Step 3: performing feature analysis on the network structure to obtain a hotspot dynamic analysis result and outputting the result.

[0095] Preferably, the preset feature set comprises a text node degree feature, an intermediary centrality feature, a text information entropy feature and a text burstiness feature.

[0096] In this embodiment, the mining samples comprise academic texts and policy texts related to the Yellow River Basin.

[0097] In this embodiment, based on the text node degree, the intermediary centrality, the text information entropy and the text burstiness, the text mining technology and the Citespace software can be used to effectively analyze the network structure features of information entities of the academic and policy system in the upper, middle and lower reaches of the Yellow River.

[0098] The above technical solution has the beneficial effect that the hotspot dynamics of different basins are determined by integrating and analyzing the academic and policy characteristics from multiple dimensions, thereby improving the efficiency of evolution research.

[0099] Embodiment 2

[0100] Based on the basis of Embodiment 1, Step 2: pre-analyzing the mining samples based on a preset feature set to determine the network structure of information entities of the policy and academic system in the Yellow River Basin, comprises the following steps of:

[0101] determining the text node degree of academic articles and policy articles in the mining samples, and further determining the influence of different nodes;

[0102] determining the intermediary centrality of the network nodes corresponding to the academic articles and policy articles in the mining samples, and further determining the application scenarios of different articles and the control ability of the articles on the keyword network based on the intermediary centrality;

[0103] determining the text information entropy of the academic articles and policy articles in the mining samples, and further determining the associated information of high-frequency keywords;

[0104] Determine the text burst characteristics of the mining sample of academic articles and policy articles, and then determine the research front hot spot of the Yellow River Basin;

[0105] Based on all the determination results, obtain the network structure of the Yellow River Basin policy and academic system information entity;

[0106] The mining sample includes academic articles and policy articles.

[0107] In this embodiment, the calculation formula of the text node degree is as follows:

[0108]

[0109] Wherein, N is the number of network nodes corresponding to the text, k(j, i) is the edge between the network node i corresponding to the text and the network node j corresponding to the text, then K i is the degree of the network node i corresponding to the text, when the edge exists, k(j, i) = 1, otherwise k(j, i) = 0, the degree of a certain text node refers to the number of edges connected to the node by other text nodes. The greater the degree of a node, the greater the role and influence of the node.

[0110] In this embodiment, the calculation formula of the intermediate centrality C(ni) of the network node i is as follows:

[0111]

[0112] Wherein, the number of shortcuts from the network node j corresponding to the text to the network node k is g jk , the number of shortcut paths from the network node j corresponding to the text to the network node k passing through the network node i corresponding to the text is g jk (ni), wherein the control ability of a certain node in the network is an important aspect concerned in the text relationship mining, which is proportional to the number of shortest path edges passing through the node.

[0113] In this embodiment, the calculation formula of the text information entropy is as follows:

[0114]

[0115] Wherein, S i0 represents the information entropy of the first article text keyword i0, p j0|i0 is the probability of the simultaneous appearance of the text keywords i0 and j0, p i0 is the probability of the appearance of the text keyword i0, wherein, in order to measure the connection ability of a certain keyword i in the text network, the other keywords associated with it need to be concerned, the information entropy S iThe larger the information carried by the keywords in the text, the greater the diversity of the associated information, and the more it can be integrated and applied to different contexts and fields. The information entropy and intermediate centrality of academic and policy texts on the upper, middle and lower reaches of the Yellow River were calculated to study the diversity of their application scenarios and their control ability over the keyword network. For academic research, the information entropy of the middle reaches of the Yellow River was the highest, at 3.7484; for policy research, the information entropy of the lower reaches of the Yellow River was the highest, at 3.1893. The keywords with high intermediate centrality in academic research and policy research were both in the middle reaches of the Yellow River, with intermediate centrality values of 1.05 and 0.96, respectively.

[0116] The beneficial effects of the above technical solutions are: by determining and calculating different features respectively, an effective construction basis is provided for constructing a network structure, and the efficiency of evolution research is indirectly improved.

[0117] Embodiment 3:

[0118] Based on the basis of Embodiment 2, the text burst feature of the mining sample is determined, including:

[0119] The sample publication time domain of the academic articles in the mining sample is obtained, and based on the sample publication time domain, the policy article acquisition time domain is determined;

[0120] The policy articles between the acquisition time domains are obtained;

[0121] All sample keywords of the academic articles and all policy keywords of the policy articles in the Yellow River basin are determined;

[0122] Based on the timestamp, the corresponding sample keywords and policy keywords at the same time are burst processed, and based on the burst processing result, the burst time length of the associated keywords of the upper reaches of the Yellow River, the middle reaches of the Yellow River and the lower reaches of the Yellow River is sorted;

[0123] At the same time, based on the sorting result, the continuity of the burst phenomenon of different associated keywords is calibrated, and then the text burst feature is determined.

[0124] In this embodiment, the changes in the number of literature and policy articles in the upper, middle and lower reaches of the Yellow River in the time series can directly reflect the dynamic evolution of academic and policy research in the upper, middle and lower reaches of the Yellow River. Since there was a literature publication interruption from 1967 to 1984, and as of 2021, the earliest policy article collected by CNKI was in 2000, only the academic literature in the upper, middle and lower reaches of the Yellow River from 1985 to 2021 and the policy articles from 2000 to 2021 were analyzed in the time series.

[0125] The research found that the number of articles in the Yellow River Basin was small from 1985 to 1995, and academic research was still in its infancy during this period. After the late 1990s, the number of articles in the upper, middle and lower reaches of the Yellow River increased significantly, with the upper reaches of the Yellow River having more articles than the middle reaches. Overall, the lower reaches of the Yellow River have the most articles. From 2015 to 2018, academic research in the Yellow River Basin developed slowly, and in 2020, there was a peak in research since 2015, and research in the middle reaches of the Yellow River increased significantly. For policy articles in the Yellow River Basin, the number of policy articles in the lower reaches of the Yellow River is slightly larger than that in the upper reaches of the Yellow River, and the policy research in the middle reaches of the Yellow River is relatively lacking. In the period from 2000 to 2010, the lower reaches of the Yellow River had a policy research peak in 2007, with 162 policy articles. The academic research in the lower reaches of the Yellow River had a peak in 2006, with 180 academic articles. In addition, from the trends of academic and policy articles in the upper, middle and lower reaches of the Yellow River, we can see that although the academic and policy research in the upper, middle and lower reaches of the Yellow River is different, the trends are similar, and the academic research in the upper, middle and lower reaches of the Yellow River generally precedes the policy research.

[0126] In the burst research of keywords in the upper reaches of the Yellow River, as shown in the table below, the keywords are sorted according to the length of the burst time, in order of hydropower station, Longyangxia Hydropower Station, Yellow River middle and upper reaches, climate change, water resources, Ning-Meng River section, ecological environment, Maqu, runoff, Qingtongxia, high-quality development, etc. Among them, the hydropower station has the longest burst duration of 17 years. In addition, high-quality development burst in 2019 and continued until 2021. We believe that this burst is continuous, so we can say that the keyword of high-quality development has the ability to continue to be a research hotspot in the coming years and is a technology branch worth paying attention to.

[0127] In the burst research of keywords in the upper, middle and lower reaches of the Yellow River, the keywords are sorted according to the length of the burst time. The academic keywords in the upper reaches of the Yellow River are sorted by burst time in order of hydropower station, Longyangxia Hydropower Station, Yellow River middle and upper reaches, climate change, water resources, Ning-Meng River section, ecological environment, Maqu, runoff, Qingtongxia, high-quality development, etc. Among them, the hydropower station has the longest burst duration of 17 years. The policy keywords in the upper reaches of the Yellow River are sorted by burst time in order of economic circle, Yellow River Basin, artificial rain enhancement, water resources development, Linxia City, China Power Investment Group, ecological protection, etc. Among them, the economic circle has the longest burst time of 12 years. By combining the burst intensity of academic and policy keywords in the upper reaches of the Yellow River, we can see that the burst intensity of high-quality development is the highest in the academic field of the upper reaches of the Yellow River, with a burst intensity of 30.36. In the policy field of the upper reaches of the Yellow River, the Yellow River Basin and ecological protection have high burst intensity, with a burst intensity of 7.46 and 6.13, respectively. It can be seen that high-quality development and ecological protection in the Yellow River Basin have received extensive attention in academic and policy research in the upper reaches of the Yellow River and have become a research frontier with great influence.

[0128] The academic burst intensity of the middle reaches of the Yellow River in order of burst time is climate change, sediment discharge, river dragon interval, small watershed management, and Loess Plateau, etc.; the policy keywords of the middle reaches of the Yellow River are in turn the Yellow River basin, hydrological station, Wuding River, Loess Plateau, and dike dam engineering, etc. Small watershed management, sediment discharge, and soil and water conservation have strong burst intensity in the academic research of the middle reaches of the Yellow River, which are 7.7, 7.05, and 6.69 respectively; in the policy research of the middle reaches of the Yellow River, hydrological station and soil and water conservation have strong burst intensity, which are 3.56 and 3.17 respectively. It can be seen that small watershed management and soil and water conservation are the focus of the academic and policy research themes in the middle reaches of the Yellow River.

[0129] River channel erosion, river channel deposition, lower river channel, Yellow River irrigation, water and sediment regulation, Sanmenxia Reservoir, water resources, and beach area management are the keywords with longer burst time in the academic research of the lower reaches of the Yellow River, among which, cutoff, water and sediment regulation, and river channel erosion have strong burst intensity in the academic research of the lower reaches of the Yellow River, which are 53.85, 22.21, and 20.35 respectively; the Yellow River basin, water and sediment regulation, standardized dike, water quantity regulation, Yellow River water resources, Yellow River irrigation area, and Xiaolangdi Project are the keywords with longer burst time in the policy of the lower reaches of the Yellow River, among which, water quantity regulation, Yellow River water resources, Xiaolangdi Reservoir, and water-sediment relationship are the keywords with higher burst intensity in the policy research of the lower reaches of the Yellow River, which are 6.91, 4.63, 4.48, and 4.45 respectively. It can be known that cutoff, water and sediment regulation, and water quantity regulation become the focus of the lower reaches of the Yellow River. In addition, high-quality development in the academic research of the upper and lower reaches of the Yellow River bursts in 2019 and continues to 2021, which indicates that this burst is continuous, that is, high-quality development has the ability to continue to be the academic research hotspot of the upper and lower reaches of the Yellow River in the coming years, and is also a technology branch worthy of attention.

[0130] The beneficial effects of the above technical solutions are: by processing the burst phenomenon, the burst intensity is obtained, which provides technical support for the academic research hotspot, and indirectly improves the efficiency of hotspot dynamic analysis.

[0131] Embodiment 4:

[0132] Based on the basis of embodiment 3, after obtaining the policy articles between the acquisition time domain, the following steps are included:

[0133] Time series analysis is performed on the academic articles and the policy articles to determine the article research trend of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River;

[0134] The article publishing agencies of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River are recorded, and then the article publishing river distribution of each agency, the author publishing distribution of each author, and the publishing cited distribution are obtained;

[0135] Based on the article publishing river distribution, author publishing distribution, publishing cited distribution and article research trend, the research field and research object of the upper reaches of the Yellow River, the middle reaches of the Yellow River and the lower reaches of the Yellow River are determined;

[0136] The first N1 cited keywords related to academic articles in the publishing cited distribution are obtained and analyzed, and the network node number, edge number and network density of the academic literature keywords of the upper reaches of the Yellow River, the middle reaches of the Yellow River and the lower reaches of the Yellow River are obtained, and then the first cited cluster of the first cited keyword of each basin is determined;

[0137] The first N2 cited keywords related to policy articles in the publishing cited distribution are obtained and analyzed, and the network node number, edge number and network density of the policy keywords of the upper reaches of the Yellow River, the middle reaches of the Yellow River and the lower reaches of the Yellow River are obtained, and then the second cited cluster of the second cited keyword of each basin is determined;

[0138] Based on the first cited cluster and the second cited cluster, the keyword co-occurrence result of the Yellow River basin is determined;

[0139] Based on the research field, research object and keyword co-occurrence result, the corresponding research hotspots are determined.

[0140] In this embodiment, the academic article authors of the upper, middle and lower reaches of the Yellow River are analyzed, and the author publishing and cooperation graph are obtained as shown in Figure 2-1 、 2-2 and 2-3, wherein the nodes represent article authors, and the node size represents the cited condition of the authors. It is found that the main academic authors of the upper, middle and lower reaches of the Yellow River are divided into 7 clusters. The authors of the upper reaches of the Yellow River, such as Luo Qiushi, An Chahua, Lu Jun, Li Chaogun, etc., have high literature cited frequency and close cooperation, followed by the authors of Li Xungui, Wei Yining and Li Fang; the authors of the middle reaches of the Yellow River, such as Zhang Jianyun, Liu Yang, Wu Houfa, Wang Guoqing, Zhang Ze and Xu Jianhua, have high literature cited frequency; the authors of the lower reaches of the Yellow River, such as Zhang Jinliang, Luo Qiushi, Zhang Hongwu, Bai Yuchuan, Chen Cuixia and Xu Haiyu, have high cited frequency. The author cited modularity of the literature of the upper, middle and lower reaches of the Yellow River is 0.9773, 0.9452 and 0.9366 respectively, and the density is 0.0041, 0.0028 and 0.0152 respectively, so it can be seen that the research of the upper, middle and lower reaches of the Yellow River appears small-scale cooperation, but the overall cooperation is lacking

[0141] The analysis of more than 5600 literatures and more than 1500 policy documents of the upper, middle and lower reaches of the Yellow River shows that the main publishing institutions and rankings of academic literatures and policy articles are shown in Table 1 and Table 2. The research institutions with the most academic publications of the upper, middle and lower reaches of the Yellow River are respectively the CPC Wuzhong Municipal Party School, the State Key Laboratory of Hydrology and Water Resources and Water Conservancy Engineering Science of Nanjing Hydraulic Research Institute and the Yellow River Survey and Design Institute Co., Ltd.; the institutions with the most policy publications of the upper, middle and lower reaches of the Yellow River are respectively Qinghai Daily and the Yellow River Newspaper. In terms of geographical distribution, the high-yield institutions of literatures and policies of the upper reaches of the Yellow River are mainly in the western region; the academic research institutions of the middle reaches of the Yellow River are evenly distributed, while the policy research institutions of the middle reaches of the Yellow River are mainly in the central and eastern regions; the central and eastern regions are the main source areas of academic and policy publications of the lower reaches of the Yellow River.

[0142] In this embodiment, as shown in Figure 3-1 , 3-2 , 3-3 and 4-1, 4-2, 4-3, the co-occurrence and clustering analysis of keywords of academic and policy articles of the upper, middle and lower reaches of the Yellow River is performed to make the keyword network structure clearer.

[0143] In this embodiment, for example, the top N1 cited keywords are the top 3% cited keywords, and the top N2 cited keywords are the top 6% cited keywords, and the determination of the number is related to the number of corresponding articles.

[0144] In this embodiment, for academic articles: the top 3% cited keywords are analyzed, and the number of nodes of the keyword network of academic literatures of the upper, middle and lower reaches of the Yellow River is respectively 118, 334 and 493, the number of edges is respectively 128, 176 and 200, and the network density is respectively 0.0185, 0.0217 and 0.0248. The main academic keywords of the upper, middle and lower reaches of the Yellow River are divided into 7, 12 and 12 clusters according to the citation frequency, and the representative keywords of the upper reaches of the Yellow River are respectively the upper reaches of the Yellow River, the Yellow River basin, the upper and middle reaches of the Yellow River, climate change, Longyangxia Hydropower Station, runoff, high-quality development; the representative keywords of the middle reaches of the Yellow River are respectively the middle reaches of the Yellow River, the Yellow River, sediment, soil and water conservation, transportation benefit, the middle reaches of the Yellow River; and the representative keywords of the lower reaches of the Yellow River are respectively the lower reaches of the Yellow River, the Yellow River, the Yellow River, the lower reaches of the Yellow River, sediment, Xiaolangdi Reservoir, river channel erosion and deposition, flood control, etc.

[0145] For policy articles: Due to the limited number of policy articles, the top 6% cited keywords were analyzed. The number of nodes in the keyword network of policy articles in the upper, middle and lower reaches of the Yellow River was 481, 488 and 415, respectively, the number of edges was 217, 219 and 181, respectively, and the network density was 0.0205, 0.0204 and 0.0255, respectively. The main policy keywords in the upper, middle and lower reaches of the Yellow River were divided into 12, 14 and 14 clusters, respectively, according to the citation frequency. The representative keywords of the upper reaches of the Yellow River were ecological protection and construction, hydropower companies, the upper reaches of the Yellow River, the West Route of the South-to-North Water Diversion Project, the Chengdu-Chongqing Economic Zone, ecological barriers, and the lower reaches of the Yellow River, etc. The representative keywords of the middle reaches of the Yellow River were the middle reaches of the Yellow River, the construction of the dam, the concentrated source area of coarse sediment, the Central Plains, the commercial and logistics base, and the river dragon interval, etc. The lower reaches of the Yellow River were the lower reaches of the Yellow River, the water resources of the Yellow River, the river administration, the lower reaches of the Yangtze River, and the lower reaches of the river, etc.

[0146] In this embodiment, as shown in Figure 5-1 , 5-2 , 6-1, 6-2, 7-1, 7-2, the research hotspots of the upper, middle and lower reaches of the Yellow River are presented. For academic research on the upper reaches of the Yellow River, the upper reaches of the Yellow River emerged in 1955 and gained more and more attention from scholars in 1987. Although the clusters of the upper and middle reaches of the Yellow River and climate change increased in the early stage, they began to cool down in 2000 and 2007, respectively, and the research attention decreased. In addition, the multi-ethnic economic development zone in the upper reaches of the Yellow River is a key word with high intermediate centrality in the cluster of cascade hydropower stations, which has influenced the trend of the whole cluster. Compared with the upper reaches, the academic research on the middle reaches of the Yellow River has fewer issues and is relatively less active. Since 1990, small watershed and loess plateau management have become the focus of soil and water conservation in the middle reaches of the Yellow River, but the research has been weak in the later stage. Although the academic research on the lower reaches of the Yellow River is less in the early stage, it has been increasing since 1983, and the clusters of the issues of concern are showing a rapid growth trend. Water resources, river regulation, water diversion and sediment, and soil conservation have high intermediate centrality, which has an important influence on the subsequent research of the clusters of the Yellow River cutoff, sediment, flood control, etc. in the lower reaches of the Yellow River.

[0147] For policy research on the upper reaches of the Yellow River, the upper and middle reaches of the Yellow River emerged in 2000, and hydropower development became a key word with high intermediate centrality in the cluster of hydropower companies in 2004. In addition, we found that the policy research on the upper reaches of the Yellow River showed a downward trend after 2016; similar to the upper reaches, the policy research on the middle reaches of the Yellow River has been less since 2010. The policy research on the lower reaches of the Yellow River is more active than the upper and middle reaches, and has more issues. Since 2013, the policy research on the lower reaches of the Yellow River has shown a decreasing trend compared with academic research. High-quality development and ecological protection have a greater impact on the cluster of water resources of the Yellow River, and will become the focus of future research on water resources of the Yellow River.

[0148] The beneficial effects of the above technical solutions are: through the cluster analysis of academic articles and policy articles, different distribution structures are facilitated, and analysis basis is provided for research hotspots.

[0149] Embodiment 5:

[0150] Based on the basis of embodiment 2, after determining the text node degree of the academic articles and policy articles in the mining sample, it further includes:

[0151] Determine the text structure formed by the mining sample, and construct the degree structure according to the obtained text node degree;

[0152] Perform consistency analysis on the text structure and the degree structure, when the consistency analysis result meets the preset division condition, according to the text node degree, retrieve the corresponding division mode from the node-division database, and perform primary division on the mining sample;

[0153] Determine the tightness between different division blocks after the primary division, determine the division precision of the division result based on the tightness, when the division precision is lower than the preset precision, obtain the first participation parameter for determining the tightness, and simultaneously obtain the second participation parameter in the division process and the division parameter of the division mode;

[0154] Based on the first participation parameter, the second participation parameter, and the division parameter, construct a parameter matrix, and according to the parameter attribute uniformity principle, first calibrate the consistent parameters in the parameter matrix, and second calibrate the inconsistent parameters;

[0155] At the same time, determine the difference parameter of the first calibrated parameter and the second calibrated parameter based on the standard parameter of the preset precision, adjust the division mode according to the difference parameter, and perform secondary division on the mining sample according to the adjusted mode, to realize the construction of the network structure.

[0156] In this embodiment, the text structure is, for example, the same type of articles published in different flow fields, according to the standard, the text structure is constructed, and then the influence of the same type of articles based on the flow field is determined, and the structure is constructed based on the text node degree, which is determined according to the influence of different nodes in different flow fields.

[0157] In the embodiment, the consistency analysis can be based on the influence of different nodes as a benchmark to determine the influence of the articles of the corresponding type of the node on the basin, such as: the influence of different nodes is a1, and the influence of the articles of the corresponding type of the node on the basin is a1, at this time, it can be considered as consistent, which meets the preset division condition (the influence of the two is the same), and then the division mode is obtained from the node-division database (node degree, division mode) according to the text node degree, and the mining sample is initially divided, this time, the initial division is performed according to the node degree, which ensures the clarity of the division.

[0158] In the embodiment, for example: there are mining samples A1, A2, A3 and A4, after division, blocks A1 and A2 and blocks A3 and A4 are obtained, the tightness of the blocks A1 and A2 and the blocks A3 and A4 is determined, that is, the degree of connection between the two, and the tightness is determined to perform fine division, for example, the blocks A1 and A2 and the blocks A3 and A4 do not meet the preset precision, but the blocks A1 and A2 meet the preset precision.

[0159] In the embodiment, the first participation parameter is 1, 2, 3 or 4, the second participation parameter is 1, 2, 3 or 6, and the division parameter is 1, 2, 6 or 7;

[0160] The determined parameter matrix is, for example: At this time, 1 and 2 can be considered as consistent parameters, and the remaining 3, 4, 6 and 7 are considered as inconsistent parameters, for example, the standard parameters based on the preset precision are 1, 2, 3 and 9, at this time, the difference parameters {(4, 9), (6, 9), (6, 3), (6, 9), (7, 3), (7, 9)} can be used to determine, and then the division mode is adjusted to perform re-division, and A1, A2 and A3, A4 are obtained.

[0161] The beneficial effects of the above technical solutions are: by constructing two structures and performing consistency analysis, the initial division can be effectively performed, when the preset division condition is met, the division time can be saved, the division efficiency can be improved, and through subsequent analysis of the precision and adjustment of the division mode according to the difference parameters, the secondary division is realized, the rationality of the division is ensured, and the basis for subsequent construction of network structure is provided.

[0162] Embodiment 6:

[0163] Based on the basis of embodiment 4, after determining the corresponding research hotspots, the following steps are further included:

[0164] The Yellow River basin is monitored, and the difficult monitoring map and the simple monitoring map are obtained according to the position monitoring result;

[0165] acquire a first monitoring sample corresponding to the difficult monitoring map, and determine a first monitoring location point according to a location classification model, and perform map analysis on the difficult monitoring map to determine the existing must-monitoring location points, if the must-monitoring location points are missing in the first monitoring location points, acquire the point location information of the missing monitoring points, and determine the survey difficulty according to the point location information, and match the subsequent survey mode related to the survey difficulty;

[0166] continuously survey the missing monitoring points according to the subsequent survey mode, and take the continuously surveyed sample as the monitoring sample corresponding to the difficult monitoring map, and perform first analysis on all samples after continuous surveying according to a monitoring index classification model, to determine the same type of monitoring location points corresponding to the same type of monitoring index in the difficult monitoring map;

[0167] acquire a second monitoring sample corresponding to the simple monitoring map, and perform second analysis on the second monitoring sample according to the monitoring index classification model, to determine the same type of monitoring location points corresponding to the same type of monitoring index in the simple monitoring map;

[0168] determine the location distribution of the same type of monitoring location points, and set a first label, and the first label includes: a location point and the index allocation related to the location point;

[0169] split the research hotspots to obtain a first hotspot set of the upper reaches of the Yellow River, and set a first hotspot label, acquire a second hotspot set of the middle reaches of the Yellow River, and set a second hotspot label, acquire a third hotspot set of the lower reaches of the Yellow River, and set a third hotspot label;

[0170] construct the label relationship among the first label, the first hotspot label, the second hotspot label, and the third hotspot label;

[0171] when the label relationship meets the preset hotspot distribution relationship, the corresponding research hotspots remain unchanged;

[0172] when the label relationship does not meet the preset hotspot distribution relationship, expand the research hotspots based on the label information of the first label, and take the expanded research hotspots as common research hotspots with the already acquired research hotspots.

[0173] In this embodiment, location monitoring is to acquire the location distribution of the entire Yellow River basin, and then to preliminarily determine the samples of different locations through the division of monitoring difficulty. Generally, the sample quantity is small in places with large monitoring difficulty.

[0174] In this embodiment, the first monitoring sample and the second monitoring sample are both acquired historically.

[0175] In this embodiment, the location classification model and the monitoring index classification model are both pre-set.

[0176] In this embodiment, the must-monitor location points refer to the monitoring data of the locations, which affect the analysis of the research hotspots.

[0177] In this embodiment, the must-monitor location points are monitored according to the corresponding survey methods to obtain samples, which are used as auxiliary information to better research and analyze the hotspots.

[0178] In this embodiment, the same type of monitoring indicators are, for example, monitoring indicators of the Yellow River Basin itself.

[0179] In this embodiment, there may be related parameters of the same type of indicators in the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River. By determining the location points corresponding to the same type of monitoring indicators, the layout of the location points corresponding to the indicators can be effectively determined.

[0180] In this embodiment, the hotspot label includes the hotspot content obtained in the above embodiments and has a label relationship with the first label, so as to better determine whether the content related to the first label needs to be used as auxiliary information to provide more data for hotspot research and analysis.

[0181] In this embodiment, the preset hotspot distribution relationship can refer to the distribution position of the preset hotspots and the like, and the label relationship is used to determine the relationship between the determined hotspots and the hotspots obtained from the auxiliary information.

[0182] The beneficial effects of the above technical solutions are that the locations are divided into difficult and easy parts, the corresponding survey methods are used to survey the difficult-to-monitor places, the effectiveness of obtaining samples is improved, and different label relationships are established to effectively supplement the researched hotspots, thereby improving the effectiveness of hotspot research.

[0183] Embodiment 7:

[0184] Based on the basis of Embodiment 1, in step 3, the network structure is subjected to feature analysis to obtain a hotspot dynamic analysis result and output, including:

[0185] Obtain the structure parameters involved in the network structure;

[0186] Based on a feature analysis model, analyze the structure parameters to obtain a feature sequence and determine sequence observation points between the feature sequences;

[0187] Determine the sequence bias of the feature sequence according to the sequence observation points of the same sequence, and perform consistent processing on the index bias of the feature index related to the feature analysis model to determine whether the consistent processing result meets a preset consistent condition. If yes, determine the hotspot dynamic corresponding to the feature sequence based on the feature index and output.

[0188] Otherwise, according to the basin information of different basins, the sequence observation points of different sequences are classified into the same basin, and the hotspot dynamics of the same basin are obtained according to the classified results, and output.

[0189] The beneficial effects of the above technical solutions are:

[0190] Embodiment 8:

[0191] Based on the basis of embodiment 1, based on the obtained hotspot dynamic analysis result and output, it further includes:

[0192] Obtain the hotspot dynamic analysis atlas of different basins in the Yellow River basin;

[0193] Obtain the corresponding sudden impact event of different basins in the Yellow River basin, and determine the actual impact index of the sudden impact event on the hotspot dynamics based on the event impact model;

[0194] According to the actual impact index, determine the actual impact value Y1;

[0195]

[0196] Wherein, S represents the number of index of the actual impact index; d s1 represents the index value of the s1th actual impact index after standard numerical conversion, and the value range is [0, 1]; f s1 (D1, d s1 ) represents the influence coefficient of the s1th actual impact index on the corresponding hotspot dynamic distribution atlas D1 of the upper reaches of the Yellow River; h1 represents the position coefficient of the upper reaches of the Yellow River and the burst position of the sudden impact event; h2 represents the position coefficient of the middle reaches of the Yellow River and the burst position of the sudden impact event; h3 represents the position coefficient of the lower reaches of the Yellow River and the burst position of the sudden impact event; f s1 (D2, d s1 ) represents the influence coefficient of the s1th actual impact index on the corresponding hotspot dynamic distribution atlas D2 of the middle reaches of the Yellow River; f s1 (D3, d s1 ) represents the influence coefficient of the s1th actual impact index on the corresponding hotspot dynamic distribution atlas D3 of the lower reaches of the Yellow River;

[0197] When the actual impact value Y1 is greater than the preset value, from f s1 (D1, d s1 ), f s1 (D2, d s1 ) and f s1 (D3, d s1screening the atlas with the influence coefficient greater than the preset coefficient, and adjusting the corresponding screened atlas based on the sudden influence event to obtain a new atlas, which is output;

[0198] When the actual influence value Y1 is less than the preset value, the corresponding actual influence index is regarded as an invalid index, and the original output atlas content remains unchanged.

[0199] In this embodiment, the dynamic analysis atlas is determined in the foregoing embodiment, and the image influence event is damage caused by human activities or sudden disasters.

[0200] In this embodiment, the actual influence index, such as the environmental index, the water resource index, and the water flow direction index, is calculated.

[0201] In this embodiment, the influence coefficient ranges from 0 to 1, the influence of the actual influence index on different basins of the Yellow River is more serious, and the corresponding influence coefficient is greater. If there is no influence, the corresponding influence coefficient is 0.

[0202] The beneficial effects of the above technical solutions are: by determining the actual influence caused by the sudden influence event, it is determined whether to change the original atlas. In the case of needing to change, the influence coefficient greater than the preset coefficient is obtained, and the adjustment is performed. This not only ensures the rationality of the atlas adjustment, but also avoids information loss, and indirectly provides effective basic content for hotspot research.

[0203] Embodiment 9:

[0204] Based on the basis of Embodiment 1, step 1: in the process of retrieving the mining sample related to the policy and academic research of the Yellow River basin from the preset database, the process includes:

[0205] retrieving a first sample related to the Yellow River basin theme from the preset database;

[0206] performing basin correlation analysis on each academic article in the first sample, retaining the academic articles with a correlation degree greater than a first preset degree, deleting the academic articles with a correlation degree less than a second preset degree, and retaining the academic articles with a correlation degree between the first preset degree and the second preset degree;

[0207] extracting the academic keywords in each academic article to be retained, determining a first number of the academic keywords, determining the research keywords in the academic keywords, and determining a second number of the research keywords;

[0208] establishing a first academic atlas of the research keywords and a second academic atlas of the academic keywords, and determining an academic overlap degree of the first academic atlas based on the second academic atlas;

[0209]

[0210] wherein, F1 represents the academic overlap degree; n1 represents the number of research keywords; n2 represents the number of remaining keywords after the research keywords are removed from the academic keywords; n3 represents the knowledge line related to each research keyword; n4 represents the knowledge line related to each remaining keyword; A2 i1,j1 represents the knowledge proportion of the j1th knowledge line related to the i1th research keyword based on the corresponding academic article; A1 i1 represents the knowledge proportion of the i1th research keyword based on the corresponding academic article; A3 i2 represents the knowledge proportion of the i2th remaining keyword based on the corresponding academic article; A4 i2,j2 represents the knowledge proportion of the j2th knowledge line related to the i2th remaining keyword based on the corresponding academic article;

[0211] When the number ratio of the second number to the second number is greater than a preset ratio or the academic overlap degree is greater than a preset overlap degree, a retention operation is performed on the corresponding academic article to be retained;

[0212] Based on the retained academic article sample and the policy article obtained in the corresponding time domain, a mining sample is constituted.

[0213] In this embodiment, the first sample, such as in the CNKI advanced search interface, respectively selects the themes “Upper Reaches of the Yellow River”, “Middle Reaches of the Yellow River”, and “Lower Reaches of the Yellow River”, and the time span is from October 1949 to July 2021, and the search is performed, irrelevant theme documents are removed, and batch export is performed according to the category of academic journals and newspapers. Among them, the academic literature with the theme of “Upper Reaches of the Yellow River” is more than 1800, and the newspapers are more than 700; the academic literature with the theme of “Middle Reaches of the Yellow River” is 1000, and the newspapers are more than 100; the academic literature with the theme of “Lower Reaches of the Yellow River” is more than 3800, and the newspapers are more than 700, and then the mining sample is screened from the first sample.

[0214] In this embodiment, the basin correlation analysis is also the correlation analysis of the upper reaches of the Yellow River, the middle reaches of the Yellow River, and the lower reaches of the Yellow River.

[0215] In this embodiment, the first preset degree and the second preset degree are both preset.

[0216] In this embodiment, the number of academic keywords is not equal to the number of research keywords, and the number of academic keywords is greater than the number of research keywords.

[0217] The above technical scheme has the beneficial effects that: the related articles are reserved by performing the correlation degree analysis, and whether the article to be reserved is reserved is determined by calculating the academic overlap degree, thereby guaranteeing the effectiveness of the mining sample and providing effective data basis for dynamic analysis of research hotspots.

[0218] Obviously, various modifications and changes can be made to the present application by those skilled in the art without departing from the spirit and scope of the application. Accordingly, it is intended that the present application embrace all such modifications and changes as fall within the scope of the claims and their equivalents.

Claims

1. A method for analyzing the dynamics of policy and academic research hotspots in the Yellow River Basin, characterized in that, include: Step 1: Retrieve mining samples related to Yellow River Basin policies and academic research from the pre-set database; Step 2: Based on a preset feature set, perform pre-analysis on the mined samples to determine the network structure of the Yellow River Basin policy and academic system information entities; Step 3: Perform feature analysis on the network structure, obtain dynamic hotspot analysis results, and output them; This also includes: Location monitoring was conducted on the Yellow River basin, and based on the location monitoring results, difficult monitoring maps and simple monitoring maps were obtained; The system acquires the first monitoring sample corresponding to the historical difficult monitoring map, and determines the first monitoring location point according to the location classification model. At the same time, it performs map analysis on the difficult monitoring map to determine the necessary monitoring location points. If the necessary monitoring location points are missing from the first monitoring location points, it acquires the location information of the missing monitoring points, determines the survey difficulty based on the location information, and matches the subsequent survey methods related to the survey difficulty. The missing monitoring points were continuously surveyed according to the subsequent survey method and used as monitoring samples corresponding to the difficult monitoring map. According to the monitoring index classification model, all samples after continuous survey were analyzed to determine the monitoring location points of the same type of monitoring index in the difficult monitoring map. Obtain the second monitoring sample corresponding to the historical simplified monitoring map, and perform a second analysis on the second monitoring sample according to the monitoring indicator classification model to determine the same type of monitoring location points corresponding to the same type of monitoring indicators in the simplified monitoring map; Determine the location distribution of similar monitoring locations and set a first label, wherein the first label includes: the location and the index allocation of the location; The research hotspots are divided to obtain the first hotspot set in the upper reaches of the Yellow River and set the first hotspot label; the second hotspot set in the middle reaches of the Yellow River and set the second hotspot label; and the third hotspot set in the lower reaches of the Yellow River and set the third hotspot label. Construct the tag relationships among the first tag, the first hotspot tag, the second hotspot tag, and the third hotspot tag; When the label relationship satisfies the preset hotspot distribution relationship, the corresponding research hotspot remains unchanged; When the label relationship does not satisfy the preset hotspot distribution relationship, the research hotspots are expanded based on the label information of the first label, and are used as common research hotspots together with the already acquired research hotspots.

2. The hotspot dynamic analysis method as described in claim 1, characterized in that, The preset feature set includes: text node degree feature, betweenness centrality feature, text information entropy feature, and text burst feature.

3. The hotspot dynamic analysis method as described in claim 1, characterized in that, Step 2: Based on a preset feature set, perform pre-analysis on the mined samples to determine the network structure of the Yellow River Basin policy and academic system information entities, including: The degree of text nodes in academic and policy articles in the mined sample is determined, and then the influence of different nodes is determined. The intermediation centrality of the network nodes corresponding to academic articles and policy articles in the mining sample is determined. Based on the intermediation centrality, the application scenarios of different articles and the control ability of the keyword network are determined. The textual information entropy of academic and policy articles in the mined sample is determined, thereby identifying the association information of high-frequency keywords; The textual burst characteristics of academic and policy articles in the mined samples were determined, thereby identifying the research frontiers and hot topics in the Yellow River Basin; Based on all the determined results, the network structure of the information entities of the Yellow River Basin policy and academic system is obtained; The samples being mined include academic articles and policy articles.

4. The hotspot dynamic analysis method as described in claim 3, characterized in that, Determining the textual burst features of the mined samples includes: Obtain the sample publication time domain of academic articles in the mining sample, and determine the policy article acquisition time domain based on the sample publication time domain; Retrieve policy articles from the aforementioned time domains; Identify all sample keywords from academic articles and all policy keywords from policy articles in the Yellow River Basin. Based on timestamps, burst processing is performed on sample keywords and policy keywords corresponding to the same time, and the burst processing results are used to sort the burst time of the associated keywords in the upper reaches, middle reaches and lower reaches of the Yellow River. Meanwhile, based on the ranking results, the continuity of the emergence of different related keywords is marked, thereby determining the text emergence characteristics.

5. The hotspot dynamic analysis method as described in claim 4, characterized in that, After obtaining the policy articles within the aforementioned time domain, the process includes: By conducting time series analysis on the academic articles and the policy articles, the research trends of the articles in the upper, middle and lower reaches of the Yellow River can be determined. Record the institutions that publish articles in the upper, middle and lower reaches of the Yellow River, and then obtain the river distribution of articles published by each institution, the author publication distribution of each author, and the citation distribution of each article. Based on the distribution of published rivers, authors' publications, citations, and research trends of the articles, the research areas and objects for the upper, middle, and lower reaches of the Yellow River are determined. The top N1 cited keywords related to academic articles in the published citation distribution are obtained and analyzed. The number of network nodes, the number of edges, and the network density of academic literature keywords in the upper, middle, and lower reaches of the Yellow River are obtained, and then the first citation cluster of the first cited keyword in each basin is determined. The top N2 cited keywords related to policy articles in the published citation distribution are obtained and analyzed. The number of network nodes, the number of edges, and the network density of policy keywords in the upper, middle, and lower reaches of the Yellow River are obtained, and then the second citation cluster of the second cited keyword in each basin is determined. Based on the first cited cluster and the second cited cluster, the keyword co-occurrence results of the Yellow River Basin are determined; Based on the research fields, research subjects, and keyword co-occurrence results, the corresponding research hotspots were identified.

6. The hotspot dynamic analysis method as described in claim 3, characterized in that, After determining the text node degree of academic and policy articles in the mined sample, the process also includes: The text structure of the mined samples is determined, and a degree structure is constructed based on the obtained text node degrees. A consistency analysis is performed on the text structure and degree structure. When the consistency analysis result meets the preset partitioning conditions, the corresponding partitioning method is retrieved from the node-partitioning database according to the text node degree to perform the initial partitioning of the mining sample. After the initial division, the density between different partitioned blocks is determined, and the division accuracy of the division result is determined based on the density. When the division accuracy is lower than the preset accuracy, the first participation parameter for determining the density is obtained. At the same time, the second participation parameter in the division process and the division parameter of the division method are obtained. Based on the first participating parameter, the second participating parameter, and the partitioning parameter, a parameter matrix is ​​constructed. In accordance with the principle of unified parameter attributes, the consistent parameters in the parameter matrix are first calibrated, and the inconsistent parameters are second calibrated. Simultaneously, the difference parameters between the first calibration parameters and the second calibration parameters based on the standard parameters of preset accuracy are determined. The division method is adjusted according to the difference parameters, and the mined samples are divided a second time according to the adjusted method to realize the construction of the network structure.

7. The hotspot dynamic analysis method as described in claim 1, characterized in that, Step 3: Perform feature analysis on the network structure, obtain dynamic hotspot analysis results, and output them, including: Obtain the structural parameters involved in the network structure; Based on the feature analysis model, the structural parameters are analyzed to obtain feature sequences, and the sequence observation points between the feature sequences are determined. Based on the sequence observation points of the same sequence, determine the sequence bias of the feature sequence and perform consistency processing with the indicator bias of the feature indicators related to the feature analysis model. Determine whether the consistency processing result meets the preset consistency conditions. If so, determine the hotspot dynamics corresponding to the feature sequence based on the feature indicators and output them. Otherwise, based on the watershed information of different watersheds, the sequence observation points of different sequences are classified into the same watershed, and the hotspot dynamics of the same watershed are obtained and output according to the classification results.

8. The hotspot dynamic analysis method as described in claim 7, characterized in that, Acquire and output the results of dynamic hotspot analysis, including: Obtain dynamic analysis maps of hotspots in different basins of the Yellow River Basin; The sudden impact events corresponding to different basins in the Yellow River Basin are obtained, and based on the event impact model, the actual impact indicators of the sudden impact events on the hotspot dynamics are determined. Based on the actual impact indicators, determine the actual impact value Y1; Where S represents the number of indicators with actual impact; d s1 f represents the standard numerical value of the s1th actual impact indicator, and its value range is [0, 1]; s1 (D1,d s1 ) represents the influence coefficient of the s1th actual impact indicator on the dynamic distribution map D1 of the hotspots in the upper reaches of the Yellow River; h1 represents the position coefficient of the upper reaches of the Yellow River and the location of the sudden impact event; h2 represents the position coefficient of the middle reaches of the Yellow River and the location of the sudden impact event; h3 represents the position coefficient of the lower reaches of the Yellow River and the location of the sudden impact event; f s1 (D2,d s1 ) represents the influence coefficient of the s1th actual impact indicator on the corresponding hotspot dynamic distribution map D2 in the middle reaches of the Yellow River; f s1 (D3,d s1 ) represents the influence coefficient of the s1th actual impact indicator on the dynamic distribution map D3 of the hotspots in the lower reaches of the Yellow River; When the actual impact value Y1 is greater than the preset value, from f s1 (D1,d s1 ), f s1 (D2,d s1 ) and f s1 (D3,d s1 The system filters out graphs with an impact coefficient greater than a preset coefficient, and adjusts the corresponding graphs based on the sudden impact event to obtain new graphs, which are then output. When the actual impact value Y1 is less than the preset value, the corresponding actual impact index is regarded as an invalid index, and the original output spectrum content remains unchanged.

9. The hotspot dynamic analysis method as described in claim 1, characterized in that, Step 1: The process of retrieving samples related to Yellow River Basin policies and academic research from the pre-set database includes: Retrieve the first sample related to the Yellow River Basin from the pre-defined database; For each academic article in the first sample, a watershed correlation analysis is performed, and academic articles with a correlation degree greater than the first preset degree are retained, while academic articles with a correlation degree less than the second preset degree are deleted. At the same time, academic articles with a correlation degree between the first preset degree and the second preset degree are to be retained. Extract academic keywords from each academic article to be retained, determine the first number of academic keywords, and simultaneously determine the research keywords among the academic keywords, and determine the second number of research keywords; Establish a first academic map of the research keywords and a second academic map of the academic keywords, and determine the academic overlap of the first academic map based on the second academic map; Where F1 represents academic overlap; n1 represents the number of research keywords; n2 represents the number of remaining keywords after removing the research keywords from the academic keywords; n3 represents the knowledge line related to each research keyword; n4 represents the knowledge line related to each remaining keyword; A2 i1,j1 A1 represents the knowledge percentage of the j1th knowledge line related to the i1th research keyword based on the corresponding academic article; i1 This indicates the knowledge percentage of the i-th research keyword based on the corresponding academic article; A3 i2 This indicates the percentage of knowledge related to the i2th remaining keyword based on the corresponding academic article; A4 i2,j2 This represents the percentage of knowledge related to the j2th knowledge line based on the corresponding academic article, which is related to the i2th remaining keyword. When the ratio of the second quantity to the second quantity is greater than a preset ratio or the academic overlap is greater than a preset overlap, the corresponding academic articles to be retained are retained. The mining sample is composed of the retained academic article sample and the policy articles obtained in the corresponding time domain.

Images