Intelligent screening and matching system and method for intelligent media editing information based on data analysis

Abstract

The invention relates to the technical field of data query, and discloses an intelligent screening and matching system and method for intelligent media editing information based on data analysis. According to the method, historical intelligent media editing information data in a network are collected, a historical intelligent media editing information data sample set is constructed, the historical intelligent media editing information data in the constructed historical intelligent media editing information data sample set are processed, comprehensive index extraction is carried out on the processed historical intelligent media editing information data, and the historical intelligent media editing information data are obtained. Based on the extracted intelligent media editing information screening factor training set and historical intelligent media editing information data, intelligent screening factors of the intelligent media editing information are determined through screening factor expansion and an intelligent screening algorithm; and finally, based on the determined intelligent screening factor of the intelligent media editing information, screening and matching are carried out on the intelligent media editing information data collected in real time, and the accuracy of intelligent screening and matching of the intelligent media editing information is improved.

Description

Technical Field

[0001] The present invention relates to the field of data query technology, and in particular to a system and method for intelligently screening and matching smart media editing information based on data analysis. Background Art

[0002] As media data continues to increase, the analysis and screening process of media data is becoming more and more complicated. Traditional data screening methods have low screening accuracy and consume too much time and resources, which has a huge impact on the final decision of media editing information.

[0003] The existing open patent application CN113449160A, this method uses preset character types and preset types to filter the stored data information to obtain matching target data items, and at the same time sorts the data values ​​contained in the obtained target data items, and matches the sorting results with the user input filtering conditions. If they match, the data information is filtered according to the filtering conditions and the sorting results of the data values ​​of the matching target data items to obtain the corresponding target data information. However, due to the preset data types, it is impossible to comprehensively classify and summarize the data types, and the filtering and matching is only performed by character type filtering, which lacks matching accuracy and has certain limitations. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the shortcomings of the prior art, the present invention provides a smart media editing information intelligent screening and matching system and method based on data analysis, which has the advantages of accuracy, real-time and high efficiency, and solves the problem that the traditional data screening method has low screening accuracy and excessive consumption of time and resources.

[0006] (II) Technical solution

[0007] In order to solve the technical problems that the above-mentioned traditional data screening method has low screening accuracy and consumes too much time and resources, the present invention provides the following technical solutions:

[0008] The present invention discloses a method for intelligently screening and matching smart media editing information based on data analysis, which specifically comprises the following steps:

[0009] S1. Collect historical intelligent media editing information data on the Internet and construct a sample set of historical intelligent media editing information data;

[0010] S2. Processing the historical smart media editing information data in the constructed historical smart media editing information data sample set to obtain processed historical smart media editing information data;

[0011] S3, extracting comprehensive indicators from the processed historical smart media editing information data to obtain a refined smart media editing information screening factor training set;

[0012] S4. Based on the refined smart media editing information screening factor training set and historical smart media editing information data, the smart media editing information screening factors are determined through screening factor expansion and smart screening algorithms;

[0013] S5. Filter and match the smart media editing information data collected in real time based on the determined intelligent screening factors of the smart media editing information.

[0014] The present invention collects historical smart media editing information data in the network and constructs a sample set of historical smart media editing information data. At the same time, the historical smart media editing information data in the constructed sample set of historical smart media editing information data is processed, and comprehensive indicators are refined for the processed historical smart media editing information data. At the same time, based on the refined smart media editing information screening factor training set and the historical smart media editing information data, the intelligent screening factors of the smart media editing information are determined through screening factor expansion and intelligent screening algorithms. Finally, based on the determined intelligent screening factors of the smart media editing information, the smart media editing information data collected in real time are screened and matched, thereby improving the accuracy of intelligent screening and matching of the smart media editing information.

[0015] Preferably, the processing of the historical smart media editing information data in the constructed historical smart media editing information data sample set to obtain the processed historical smart media editing information data comprises the following steps:

[0016] S21. Randomly select a set of processed historical intelligent media edited information data from the sample set, extract all information keywords, and construct an information keyword set B;

[0017] S22. For each information keyword in the information keyword set B, calculate the probability that the information keyword becomes the subject of the processed historical intelligent media edited information data;

[0018]

[0019] Wherein, Q(b|g) represents the probability that the information keyword b in the information keyword set B becomes the theme of the historical intelligent media edited information data after the g-th group is processed, and g(b) represents the number of times the information keyword b appears in the historical intelligent media edited information data after the g-th group is processed;

[0020] S23. Based on the calculated probability of each information keyword becoming a topic, determine the processed historical smart media editing information data.

[0021] Preferably, the determining of the processed historical intelligent media edited information data based on the calculated probability of each information keyword becoming a topic comprises the following steps:

[0022] The formula for calculating the weight of information keywords is as follows:

[0023] W b =a×Q(b|g)+c×local b ;

[0024] Among them, W b represents the weight of the news keyword b, a represents the probability parameter, c represents the position parameter, local b Indicates the position of news keyword b in the historical smart media edited news data;

[0025] A weight threshold of information keywords is set, information keywords whose weight is greater than or equal to the weight threshold are summarized, and the summarized information keywords are set as the processed historical smart media editing information data.

[0026] The present invention selects historical smart media edited information data and extracts all information keywords. At the same time, based on the calculated probability of keywords becoming topics and the weights of keywords, the processed historical smart media edited information data is determined, thereby improving the rationality of determining the historical smart media edited information data.

[0027] Preferably, the step of extracting comprehensive indicators from the processed historical intelligent media editing information data to obtain the extracted intelligent media editing information screening factor training set comprises the following steps:

[0028] S31, constructing an indicator matrix of the processed historical intelligent media editing information data;

[0029] Set each information keyword in each group of processed historical intelligent media edited information data as an indicator;

[0030] Furthermore, an indicator matrix of size n×m is constructed based on the m indicators contained in the n groups of processed historical intelligent media editing information data;

[0031] Set x ij is the jth indicator in the i-th group of processed historical intelligent media editing information data, where i∈{1,2,...,n}; j∈{1,2,...m};

[0032] S32, normalizing the indicators in the constructed indicator matrix;

[0033] Furthermore, the indicators are normalized according to the positive indicator calculation formula and the negative indicator calculation formula;

[0034] Positive indicator calculation formula:

[0035]

[0036] Negative indicator calculation formula:

[0037]

[0038] in, represents the jth positive indicator in the output of the i-th group of processed historical intelligent media editing information data, x min Indicates the minimum value of the index in the historical intelligent media editing information data after processing, x max Indicates the maximum value of the index in the processed historical intelligent media editing information data. represents the jth negative indicator in the output of the i-th group of processed historical intelligent media editing information data, x i ' j represents the jth indicator in the normalized i-th group of processed historical intelligent media editing information data;

[0039] S33. Determine the indicator weights by entropy weight method;

[0040] S34. Comprehensive indicators are refined based on the normalized indicators and their weights to obtain a refined intelligent media editor information screening factor training set.

[0041] Preferably, determining the indicator weights by the entropy weight method comprises the following steps:

[0042] S331. The formula for determining the index weight using the entropy weight method is:

[0043]

[0044] Among them, P ij is the weight value of the jth indicator in the historical intelligent media editing information data after normalization of the i-th group;

[0045] Calculate the entropy value e of the jth indicator j , the formula is:

[0046]

[0047] S332. Assign weights to the indicators and obtain the weights W of each indicator j ;

[0048] The indicator weight calculation formula is as follows:

[0049]

[0050] Among them, W j is the indicator weight of the jth indicator, h jRepresents the utility value of the j-th indicator.

[0051] Preferably, the step of refining comprehensive indicators based on normalized indicators and their weights to obtain a refined intelligent media editorial information screening factor training set comprises the following steps:

[0052] S341, standardizing the normalized indicators;

[0053]

[0054] Among them, z ij Indicates the standardized indicators;

[0055] S342. Extract comprehensive indicators based on standardized indicators and their weights;

[0056] Set the maximum value of each indicator data as the positive ideal solution, and the minimum value of each indicator data as the negative ideal solution;

[0057] Calculate the distance between each standardized index and the positive ideal solution and the negative ideal solution;

[0058] Forward distance d + = weight of each indicator × the difference between the positive ideal solution and the standardized indicators;

[0059] Negative distance d - = weight of each indicator × the difference between each indicator after standardization and the negative ideal solution;

[0060] The comprehensive index extraction formula is as follows:

[0061]

[0062] Among them, f i It represents the relative score of each indicator of the historical intelligent media editing information data after the i-th group is processed. It represents the negative distance of each indicator in the historical intelligent media editing information data after the i-th group is processed. It represents the positive distance of each indicator in the historical intelligent media editing information data after the i-th group is processed;

[0063] A relative scoring threshold is set, and the indicators that are greater than the set scoring threshold are summarized as the refined smart media editorial information screening factor training set.

[0064] The present invention constructs an indicator matrix of processed historical smart media editing information data, normalizes the indicators in the constructed indicator matrix, determines the indicator weights by the entropy weight method, and finally refines the comprehensive indicators based on the normalized indicators and the indicator weights to obtain the refined smart media editing information screening factor training set, thereby improving the reliability of the calculation of the screening factor training set.

[0065] Preferably, the step of determining the intelligent screening factors of the intelligent media editing information based on the refined intelligent media editing information screening factor training set and the historical intelligent media editing information data through screening factor expansion and intelligent screening algorithm comprises the following steps:

[0066] S41, expanding the refined intelligent media editorial information screening factor training set;

[0067] All the data in the screening factors in the refined smart media editorial information screening factor training set are randomly shuffled and combined, and the combined factors are set as shadow factors;

[0068] Merge the shadow factors with the screening factors of the screening factor training set to obtain an expanded screening factor training set;

[0069] S42. Based on the expanded screening factor training set, the intelligent screening factors for various types of smart media editorial information are determined through intelligent screening algorithms.

[0070] Preferably, the step of determining the intelligent screening factors of various types of intelligent media editing information through an intelligent screening algorithm based on the expanded screening factor training set comprises the following steps:

[0071] S421, for the expanded screening factor training set, randomly select R samples with replacement, and build a decision tree based on the selected R samples, and set the selected R samples as samples at the root node of the decision tree;

[0072] S422, when each sample has k attributes, randomly select one attribute from the k attributes as the classification attribute of the node;

[0073] Set each classification attribute to be selected only once, and each classification will only generate two nodes;

[0074] S423, classifying each node in the decision tree according to step S422 until the sample can no longer be classified, and constructing a decision tree based on the classified nodes;

[0075] S424. Construct a random forest based on the constructed decision tree.

[0076] Preferably, constructing a random forest based on the constructed decision tree comprises the following steps:

[0077] Set the number of decision trees to be constructed, and construct the required number of decision trees according to steps S421-S423;

[0078] Summarize the constructed decision trees and construct a random forest. Set the constructed random forest as a data set, and each decision tree as a set of data in the data set.

[0079] Summarize the classification results of each decision tree in the random forest, merge the same results, and use the classification attributes as the cluster center of the current classification results. Each cluster center represents an intelligent screening factor.

[0080] Output the summarized cluster centers of each category and the historical intelligent media editing information data of each category, and save them to the dataset.

[0081] The present invention obtains a refined smart media editing information screening factor training set through expansion, and constructs multiple groups of decision trees through an intelligent screening algorithm based on the expanded screening factor training set, which are combined into a random forest. At the same time, the intelligent screening factors of historical smart media editing information data of each category are determined through the combined random forest, thereby improving the efficiency of smart media editing information data screening.

[0082] Preferably, the screening and matching of the real-time collected smart media editing information data based on the determined intelligent screening factor of the smart media editing information comprises the following steps:

[0083] The smart media editing information data collected in real time is processed in step S2 to obtain processed real-time smart media editing information data;

[0084] The obtained real-time smart media editing information data is matched and compared with the intelligent screening factors one by one, and the real-time collected smart media editing information data is screened according to the comparison results.

[0085] The present invention also discloses a smart media editing information intelligent screening and matching system based on data analysis, which is used to implement a smart media editing information intelligent screening and matching method based on data analysis. The system includes: a data acquisition module, a data processing module, a screening factor refining module, a screening factor intelligent determination module, and a screening matching module;

[0086] The data collection module is used to collect intelligent media editing information data in the network;

[0087] The data processing module is used to process the collected intelligent media editing information data to obtain processed intelligent media editing information data;

[0088] The screening factor refining module is used to perform comprehensive refining based on the processed smart media editing information data to determine the smart media editing information screening factor training set;

[0089] The screening factor intelligent determination module is used to classify and screen the smart media editing information screening factor training set to determine the smart media editing information intelligent screening factors;

[0090] The screening and matching module is used to screen and match the smart media editing information data collected in real time according to the intelligent screening factors.

[0091] (III) Beneficial effects

[0092] Compared with the prior art, the present invention provides a system and method for intelligent screening and matching of smart media editorial information based on data analysis, which has the following beneficial effects:

[0093] 1. The invention collects historical smart media editing information data on the network and constructs a sample set of historical smart media editing information data. At the same time, the historical smart media editing information data in the constructed sample set of historical smart media editing information data is processed, and comprehensive indicators are extracted from the processed historical smart media editing information data. At the same time, based on the refined smart media editing information screening factor training set and historical smart media editing information data, the intelligent screening factors of smart media editing information are determined through screening factor expansion and intelligent screening algorithms. Finally, based on the determined intelligent screening factors of smart media editing information, the real-time collected smart media editing information data is screened and matched, thereby improving the accuracy of intelligent screening and matching of smart media editing information.

[0094] 2. The invention selects historical smart media edited information data and extracts all information keywords. At the same time, based on the calculated probability of the keywords becoming themes and the weights of the keywords, it determines the processed historical smart media edited information data, thereby improving the rationality of determining the historical smart media edited information data.

[0095] 3. The invention constructs an indicator matrix of processed historical smart media editing information data, normalizes the indicators in the constructed indicator matrix, determines the indicator weights by the entropy weight method, and finally extracts comprehensive indicators based on the normalized indicators and the weights of each indicator to obtain the refined smart media editing information screening factor training set, thereby improving the reliability of the calculation of the screening factor training set.

[0096] 4. The invention obtains a refined training set of smart media editing information screening factors through expansion, and constructs multiple groups of decision trees through an intelligent screening algorithm based on the expanded screening factor training set, which are combined into a random forest. At the same time, the intelligent screening factors of each category of historical smart media editing information data are determined through the combined random forest, thereby improving the efficiency of smart media editing information data screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 It is a structural diagram of the intelligent screening and matching process of the smart media editing information of the present invention. DETAILED DESCRIPTION

[0098] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0099] Example 1

[0100] See also Figure 1 This embodiment discloses a method for intelligently screening and matching smart media editorial information based on data analysis, which specifically includes the following steps:

[0101] S1. Collect historical intelligent media editing information data on the Internet and construct a sample set of historical intelligent media editing information data;

[0102] S2. Processing the historical smart media editing information data in the constructed historical smart media editing information data sample set to obtain processed historical smart media editing information data;

[0103] S3, extracting comprehensive indicators from the processed historical smart media editing information data to obtain a refined smart media editing information screening factor training set;

[0104] S4. Based on the refined smart media editing information screening factor training set and historical smart media editing information data, the smart media editing information screening factors are determined through screening factor expansion and smart screening algorithms;

[0105] S5. Screening and matching the smart media editing information data collected in real time based on the determined smart media editing information intelligent screening factors;

[0106] For further information, see Figure 1 , processing the historical smart media editing information data in the constructed historical smart media editing information data sample set to obtain the processed historical smart media editing information data includes the following steps:

[0107] S21. Randomly select a set of processed historical intelligent media edited information data from the sample set, extract all information keywords, and construct an information keyword set B;

[0108] S22. For each information keyword in the information keyword set B, calculate the probability that the information keyword becomes the subject of the processed historical intelligent media edited information data;

[0109]

[0110] Wherein, Q(b|g) represents the probability that the information keyword b in the information keyword set B becomes the theme of the historical intelligent media edited information data after the g-th group is processed, and g(b) represents the number of times the information keyword b appears in the historical intelligent media edited information data after the g-th group is processed;

[0111] S23, based on the calculated probability of each information keyword becoming a topic, determining the processed historical intelligent media edited information data;

[0112] The formula for calculating the weight of information keywords is as follows:

[0113] W b =a×Q(b|g)+c×local b ;

[0114] Among them, W b represents the weight of the news keyword b, a represents the probability parameter, c represents the position parameter, local b Indicates the position of news keyword b in the historical smart media edited news data, including title, subtitle, text and summary;

[0115] Setting a weight threshold of information keywords, aggregating information keywords whose weights are greater than or equal to the weight threshold, and setting the aggregated information keywords as the processed historical intelligent media edited information data;

[0116] For further information, see Figure 1 , extracting comprehensive indicators from the processed historical smart media editing information data to obtain the refined smart media editing information screening factor training set includes the following steps:

[0117] S31, constructing an indicator matrix of the processed historical intelligent media editing information data;

[0118] Set each information keyword in each group of processed historical intelligent media edited information data as an indicator;

[0119] Furthermore, an indicator matrix of size n×m is constructed based on the m indicators contained in the n groups of processed historical intelligent media editing information data;

[0120] Set x ij is the jth indicator in the i-th group of processed historical intelligent media editing information data, where i∈{1,2,...,n}; j∈{1,2,...m};

[0121] S32, normalizing the indicators in the constructed indicator matrix;

[0122] Furthermore, the indicators are normalized according to the positive indicator calculation formula and the negative indicator calculation formula;

[0123] Positive indicator calculation formula:

[0124]

[0125] Negative indicator calculation formula:

[0126]

[0127] in, represents the jth positive indicator in the output of the i-th group of processed historical intelligent media editing information data, x min Indicates the minimum value of the index in the historical intelligent media editing information data after processing, x max Indicates the maximum value of the index in the processed historical intelligent media editing information data. represents the jth negative indicator in the output of the i-th group of processed historical intelligent media editing information data, x i ' j represents the jth indicator in the normalized i-th group of processed historical intelligent media editing information data;

[0128] S33. Determine the indicator weights by entropy weight method;

[0129] S331. The formula for determining the index weight using the entropy weight method is:

[0130]

[0131] Among them, P ij is the weight value of the jth indicator in the historical intelligent media editing information data after normalization of the i-th group;

[0132] Calculate the entropy value e of the jth indicator j , the formula is:

[0133]

[0134] S332. Assign weights to the indicators and obtain the weights W of each indicator j ;

[0135] The indicator weight calculation formula is as follows:

[0136]

[0137] Among them, W j is the indicator weight of the jth indicator, h j represents the utility value of the jth indicator;

[0138] S34, refining comprehensive indicators based on normalized indicators and their weights, and obtaining a refined intelligent media editorial information screening factor training set;

[0139] S341, standardizing the normalized indicators;

[0140]

[0141] Among them, z ij Indicates the standardized indicators;

[0142] S342. Extract comprehensive indicators based on standardized indicators and their weights;

[0143] Set the maximum value of each indicator data as the positive ideal solution, and the minimum value of each indicator data as the negative ideal solution;

[0144] Calculate the distance between each standardized index and the positive ideal solution and the negative ideal solution;

[0145] Forward distance d + = weight of each indicator × the difference between the positive ideal solution and the standardized indicators;

[0146] Negative distance d - = weight of each indicator × the difference between each indicator after standardization and the negative ideal solution;

[0147] The comprehensive index extraction formula is as follows:

[0148]

[0149] Among them, f i It represents the relative score of each indicator of the historical intelligent media editing information data after the i-th group is processed. It represents the negative distance of each indicator in the historical intelligent media editing information data after the i-th group is processed. It represents the positive distance of each indicator in the historical intelligent media editing information data after the i-th group is processed;

[0150] Furthermore, a relative scoring threshold is set, and the indicators greater than the set scoring threshold are summarized as the refined intelligent media editorial information screening factor training set;

[0151] For further information, see Figure 1 Based on the refined intelligent media editing information screening factor training set and historical intelligent media editing information data, through screening factor expansion and intelligent screening algorithm, the intelligent screening factors of intelligent media editing information are determined, including the following steps:

[0152] S41, expanding the refined intelligent media editorial information screening factor training set;

[0153] All the data in the screening factors in the refined smart media editorial information screening factor training set are randomly shuffled and combined, and the combined factors are set as shadow factors;

[0154] Merge the shadow factors with the screening factors of the screening factor training set to obtain an expanded screening factor training set;

[0155] S42, based on the expanded screening factor training set, using the intelligent screening algorithm, determine the intelligent screening factors for various types of intelligent media editing information;

[0156] S421, for the expanded screening factor training set, randomly select R samples with replacement, and build a decision tree based on the selected R samples, and set the selected R samples as samples at the root node of the decision tree;

[0157] S422, when each sample has k attributes, randomly select one attribute from the k attributes as the classification attribute of the node;

[0158] Set each classification attribute to be selected only once, and each classification will only generate two nodes;

[0159] S423, classifying each node in the decision tree according to step S422 until the sample can no longer be classified, and constructing a decision tree based on the classified nodes;

[0160] S424. Construct a random forest based on the constructed decision tree.

[0161] Preferably, constructing a random forest based on the constructed decision tree comprises the following steps:

[0162] Set the number of decision trees to be constructed, and construct the required number of decision trees according to steps S421-S423;

[0163] Summarize the constructed decision trees and construct a random forest. Set the constructed random forest as a data set, and each decision tree as a set of data in the data set.

[0164] Summarize the classification results of each decision tree in the random forest, merge the same results, and use the classification attributes as the cluster center of the current classification results. Each cluster center represents an intelligent screening factor.

[0165] Output the aggregated cluster centers of each category and the historical intelligent media editing information data of each category, and save them to the data set;

[0166] For further information, see Figure 1 , based on the determined intelligent screening factors of the intelligent media editing information, screening and matching the real-time collected intelligent media editing information data includes the following steps:

[0167] The smart media editing information data collected in real time is processed in step S2 to obtain processed real-time smart media editing information data;

[0168] The obtained real-time intelligent media editing information data is matched and compared with the intelligent screening factors one by one, and the real-time collected intelligent media editing information data is screened according to the comparison results;

[0169] Example 2

[0170] See also Figure 1 , this embodiment also discloses a smart media editing information intelligent screening and matching system based on data analysis, which is used to implement the smart media editing information intelligent screening and matching method based on data analysis. The system includes: a data acquisition module, a data processing module, a screening factor refining module, a screening factor intelligent determination module and a screening matching module;

[0171] The data collection module is used to collect intelligent media editing information data in the network;

[0172] The data processing module is used to process the collected intelligent media editing information data to obtain processed intelligent media editing information data;

[0173] The screening factor refining module is used to perform comprehensive refining based on the processed smart media editing information data to determine the smart media editing information screening factor training set;

[0174] The screening factor intelligent determination module is used to classify and screen the smart media editing information screening factor training set to determine the smart media editing information intelligent screening factors;

[0175] The screening and matching module is used to screen and match the smart media editing information data collected in real time according to the intelligent screening factors.

[0176] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent screening and matching of intelligent media editorial information based on data analysis, characterized in that: The following steps are involved: S1. Collect historical intelligent media editing information data on the Internet and construct a sample set of historical intelligent media editing information data; S2. Processing the historical smart media editing information data in the constructed historical smart media editing information data sample set to obtain processed historical smart media editing information data; S3, extracting comprehensive indicators from the processed historical smart media editing information data to obtain a refined smart media editing information screening factor training set; S4. Based on the refined smart media editing information screening factor training set and historical smart media editing information data, the smart media editing information screening factors are determined through screening factor expansion and smart screening algorithms; S5. Filter and match the smart media editing information data collected in real time based on the determined intelligent screening factors of the smart media editing information.

2. According to claim 1, a method for intelligent screening and matching of intelligent media editing information based on data analysis is characterized in that: The processing of the historical intelligent media editing information data in the constructed historical intelligent media editing information data sample set to obtain the processed historical intelligent media editing information data includes the following steps: S21. Randomly select a set of processed historical intelligent media edited information data from the sample set, extract all information keywords, and construct an information keyword set B; S22. For each information keyword in the information keyword set B, calculate the probability that the information keyword becomes the subject of the processed historical intelligent media edited information data; Wherein, Q(b|g) represents the probability that the information keyword b in the information keyword set B becomes the theme of the historical intelligent media edited information data after the g-th group is processed, and g(b) represents the number of times the information keyword b appears in the historical intelligent media edited information data after the g-th group is processed; S23. Based on the calculated probability of each information keyword becoming a topic, determine the processed historical smart media editing information data.

3. According to claim 2, a method for intelligent screening and matching of intelligent media editing information based on data analysis is characterized in that: The step of determining the processed historical intelligent media edited information data based on the calculated probability of each information keyword becoming a topic includes the following steps: The formula for calculating the weight of information keywords is as follows: W b =a×Q(b|g)+c×local b ; Among them, W b represents the weight of the news keyword b, a represents the probability parameter, c represents the position parameter, local b Indicates the position of news keyword b in the historical smart media edited news data; A weight threshold of information keywords is set, information keywords whose weight is greater than or equal to the weight threshold are summarized, and the summarized information keywords are set as the processed historical smart media editing information data.

4. According to claim 1, a method for intelligent screening and matching of intelligent media editing information based on data analysis is characterized in that: The comprehensive index extraction of the processed historical intelligent media editing information data to obtain the extracted intelligent media editing information screening factor training set comprises the following steps: S31, constructing an indicator matrix of the processed historical intelligent media editing information data; S32, normalizing the indicators in the constructed indicator matrix; Normalize the indicators according to the positive indicator calculation formula and the negative indicator calculation formula; in, represents the jth positive indicator in the outputted i-th group of processed historical intelligent media editing information data, x min Indicates the minimum value of the index in the processed historical intelligent media editing information data, x max Indicates the maximum value of the index in the processed historical intelligent media editing information data. represents the jth negative indicator in the output of the i-th group of processed historical intelligent media editing information data, x' ij represents the jth indicator in the normalized i-th group of historical intelligent media edited information data; S33. Determine the indicator weights by entropy weight method; S34. Comprehensive indicators are refined based on the normalized indicators and their weights to obtain a refined intelligent media editor information screening factor training set.

5. According to claim 4, a method for intelligent screening and matching of intelligent media editing information based on data analysis is characterized in that: Determining the indicator weights by the entropy weight method comprises the following steps: S331. The formula for determining the index weight using the entropy weight method is: Among them, P ij is the weight value of the jth indicator in the historical intelligent media editing information data after normalization of the i-th group; Calculate the entropy value e of the jth indicator j , the formula is: S332. Assign weights to the indicators and obtain the weights W of each indicator j ; The indicator weight calculation formula is as follows: Among them, W j is the indicator weight of the jth indicator, h j Represents the utility value of the j-th indicator.

6. According to claim 4, a method for intelligent screening and matching of intelligent media editing information based on data analysis is characterized in that: The step of refining comprehensive indicators based on normalized indicators and their weights to obtain a refined intelligent media editorial information screening factor training set includes the following steps: S341, standardizing the normalized indicators; Among them, z ij Indicates the standardized indicators; S342. Extract comprehensive indicators based on standardized indicators and their weights; Set the maximum value of each indicator data as the positive ideal solution, and the minimum value of each indicator data as the negative ideal solution; Calculate the distance between each standardized index and the positive ideal solution and the negative ideal solution; Forward distance d + = weight of each indicator × the difference between the positive ideal solution and the standardized indicators; Negative distance d - = weight of each indicator × the difference between each indicator after standardization and the negative ideal solution; The comprehensive index extraction formula is as follows: Among them, f i It represents the relative score of each indicator of the historical intelligent media editing information data after the i-th group is processed. It represents the negative distance of each indicator in the historical intelligent media editing information data after the i-th group is processed. It represents the positive distance of each indicator in the historical intelligent media editing information data after the i-th group is processed; A relative scoring threshold is set, and the indicators that are greater than the set scoring threshold are summarized as the refined smart media editorial information screening factor training set.

7. According to claim 1, a method for intelligent screening and matching of intelligent media editing information based on data analysis is characterized in that: The method of determining the intelligent screening factors of intelligent media editing information based on the refined intelligent media editing information screening factor training set and historical intelligent media editing information data by screening factor expansion and intelligent screening algorithm includes the following steps: S41, expanding the refined intelligent media editorial information screening factor training set; S42. Based on the expanded screening factor training set, the intelligent screening factors for various types of smart media editorial information are determined through intelligent screening algorithms.

8. The method for intelligent screening and matching of intelligent media editing information based on data analysis according to claim 7 is characterized in that: Determining the intelligent screening factors of various types of intelligent media editing information through an intelligent screening algorithm based on the expanded screening factor training set includes the following steps: S421, for the expanded screening factor training set, randomly select R samples with replacement, and build a decision tree based on the selected R samples, and set the selected R samples as samples at the root node of the decision tree; S422, when each sample has k attributes, randomly select one attribute from the k attributes as the classification attribute of the node; Set each classification attribute to be selected only once, and each classification will only generate two nodes; S423, classifying each node in the decision tree according to step S422 until the sample can no longer be classified, and constructing a decision tree based on the classified nodes; S424. Construct a random forest based on the constructed decision tree. Preferably, constructing a random forest based on the constructed decision tree comprises the following steps: Set the number of decision trees to be constructed, and construct the required number of decision trees according to steps S421-S423; Summarize the constructed decision trees and construct a random forest. Set the constructed random forest as a data set, and each decision tree as a set of data in the data set. Summarize the classification results of each decision tree in the random forest, merge the same results, and use the classification attributes as the cluster center of the current classification results. Each cluster center represents an intelligent screening factor. Output the summarized cluster centers of each category and the historical intelligent media editing information data of each category, and save them to the dataset.

9. The method for intelligent screening and matching of intelligent media editing information based on data analysis according to claim 1, characterized in that: The screening and matching of the real-time collected smart media editing information data based on the determined smart media editing information intelligent screening factor includes the following steps: The smart media editing information data collected in real time is processed in step S2 to obtain processed real-time smart media editing information data; The obtained real-time smart media editing information data is matched and compared with the intelligent screening factors one by one, and the real-time collected smart media editing information data is screened according to the comparison results.

10. A system for implementing the intelligent screening and matching method of smart media editing information based on data analysis as described in any one of claims 1 to 9, characterized in that: include: Data collection module, data processing module, screening factor extraction module, screening factor intelligent determination module and screening matching module; The data collection module is used to collect intelligent media editing information data in the network; The data processing module is used to process the collected intelligent media editing information data to obtain processed intelligent media editing information data; The screening factor refining module is used to perform comprehensive refining based on the processed smart media editing information data to determine the smart media editing information screening factor training set; The screening factor intelligent determination module is used to classify and screen the smart media editing information screening factor training set to determine the smart media editing information intelligent screening factors; The screening and matching module is used to screen and match the smart media editing information data collected in real time according to the intelligent screening factors.

Images