A classification method and device for multiple historical and cultural resources

Abstract

The present invention discloses a classification method and device for multiple historical and cultural resources, including the following steps: obtaining spatial entities belonging to historical and cultural resources within a target range, generating a list of historical and cultural resource spatial entity items; extracting cultural characteristic dimensions, cultural characteristic dimension weights, and cultural attributes covered by each cultural characteristic dimension of historical and cultural resources, constructing a "historical and cultural resource-cultural attribute" relationship mapping table, and a "cultural attribute-cultural attribute" similarity matrix under each cultural dimension; calculating the similarity distance between any two historical and cultural resource spatial entities, constructing a "historical and cultural resource-historical and cultural resource" similarity distance matrix; using different clustering methods to generate different tree-like clustering structures, selecting the clustering result corresponding to the tree-like clustering structure with the largest silhouette coefficient as the classification result of the historical and cultural resources, and visually outputting it. The present invention has high accuracy and efficiency.

Description

Technical Field

[0001] The present invention relates to computer data processing technology, and in particular to a classification method and equipment for multiple historical and cultural resources. Background Art

[0002] As the protection and inheritance of historical and cultural heritage has gained attention, how to systematically classify or categorize historical and cultural resources that are widely distributed and have multicultural attributes is an especially important basic issue in this field. The existing classification or categorization methods of historical and cultural resources mainly rely on the existing experience of planners, which can be briefly summarized into two models: "time-space clustering" and "cultural clustering", focusing on representative historical and cultural resources or specific a priori cultural dimensions. That is, the current classification method relies more on subjective consciousness and is limited to a single dimension. The time-space organization logic of the spatial entity of historical and cultural resources is often ignored, so it often requires a lot of subjective deliberation and trade-offs, which are inefficient and highly subjective. Summary of the invention

[0003] In view of the problems existing in the prior art, the purpose of the present invention is to provide a classification method and equipment for multiple historical and cultural resources with high efficiency, high degree of automation, strong objectivity and high accuracy.

[0004] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0005] A classification method for multiple historical and cultural resources includes the following steps:

[0006] Step 1: Obtain the spatial entities belonging to historical and cultural resources within the target range and generate a list of historical and cultural resource spatial entity items;

[0007] Step 2: According to the list of historical and cultural resource space entity entries, retrieve the relevant data text of the historical and cultural resource space entity, and extract the cultural characteristic dimensions of the historical and cultural resources, the cultural characteristic dimension weights and the cultural attributes covered by each cultural characteristic dimension through natural language processing of the relevant data text, and finally construct a "historical and cultural resource-cultural attribute" relationship mapping table and a "cultural attribute-cultural attribute" similarity matrix under each cultural dimension;

[0008] Step 3: According to the "historical and cultural resources-cultural attributes" relationship mapping table and the "cultural attributes-cultural attributes" similarity matrix, the similarity distance between any two historical and cultural resource space entities is calculated, and the "historical and cultural resources-historical and cultural resources" similarity distance matrix is ​​constructed;

[0009] Step 4: Based on the “historical and cultural resources-historical and cultural resources” similarity distance matrix, different clustering methods are used to generate different tree clustering structures, and the tree clustering structure with the largest silhouette coefficient and the number of clusters are selected as the classification result of historical and cultural resources.

[0010] Furthermore, the method further comprises the following steps:

[0011] Step 5: Based on the “historical and cultural resources-historical and cultural resources” similarity distance matrix, draw a three-dimensional scatter plot that can present the classification results, protection importance, administrative location, and cultural and tourism popularity of the historical and cultural resource spatial entities.

[0012] Furthermore, the step 1 specifically includes:

[0013] Step 1.1: Collect relevant texts related to the protection and inheritance of historical and cultural resources, and extract the spatial entity types belonging to historical and cultural resources through natural language processing;

[0014] Step 1.2: According to the spatial entity types belonging to historical and cultural resources, obtain the spatial entities belonging to the spatial entity types of historical and cultural resources within the target range, and generate a list of historical and cultural resource spatial entity items.

[0015] Furthermore, the step 2 specifically includes:

[0016] Step 2.1, obtain several data texts related to each historical and cultural resource space entity in the historical and cultural resource space entity entry list from existing data documents, and form a related text set with all data texts related to all historical and cultural resource space entities;

[0017] Step 2.2, performing natural language processing on the relevant text set, extracting a number of cultural feature dimensions and corresponding weights related to historical and cultural resources, as well as all cultural attributes covered by each cultural feature dimension;

[0018] Step 2.3, perform natural language processing on a number of data texts related to each historical and cultural resource space entity, extract the cultural attributes of each historical and cultural resource space entity in each cultural feature dimension, and establish a "historical and cultural resource-cultural attribute" relationship mapping table, which covers the mapping relationship between each historical and cultural resource space entity and all its cultural attributes;

[0019] Step 2.4: Calculate the semantic similarity between any two cultural attributes under each cultural feature dimension to form a "cultural attribute-cultural attribute" similarity matrix under each cultural feature dimension.

[0020] Furthermore, in step 2.3, for the historical and cultural resource space entity with multiple cultural attributes under the same cultural characteristic dimension, the mapping relationship between the historical and cultural resource space entity and the different cultural attributes of the same cultural characteristic dimension is added as a mapping relationship to the "historical and cultural resources-cultural attributes" relationship mapping table, and the priority of multiple mapping relationships in the same cultural characteristic dimension is noted according to the degree of relevance.

[0021] Furthermore, the step 3 specifically includes:

[0022] Step 3.1, according to the "historical and cultural resources-cultural attributes" relationship mapping table and the "cultural attributes-cultural attributes" similarity matrix, obtain the similarity of any two historical and cultural resource space entities in each cultural feature dimension, and form all similarity data into a similarity data set;

[0023] Step 3.2: Based on the similarity dataset, measure the similarity distance between any two historical and cultural resource space entities according to the following formula:

[0024]

[0025] In the formula, d(H x ,H y ) is the historical and cultural resource space entity H x , H y The similarity distance, p is the total number of cultural feature dimensions, W i is the weight of the i-th cultural characteristic dimension, L i is the similarity between any two historical and cultural resource space entities in the similarity dataset under the i-th cultural feature dimension, σ i is the standard deviation of all data under the i-th cultural feature dimension of all historical and cultural resource space entities in the similarity dataset;

[0026] Step 3.3: Based on the similarity distances between any two historical and cultural resource space entities, construct a "historical and cultural resource-historical and cultural resource" similarity distance matrix; among which, the similarity distance between the same historical and cultural resource space entities is 0.

[0027] Furthermore, the step 4 specifically includes:

[0028] Step 4.1: clustering is performed using a plurality of clustering algorithms with different standards according to the similarity distance matrix of "historical and cultural resources-historical and cultural resources" to obtain a plurality of clustering tree structures;

[0029] Step 4.2: Calculate the silhouette coefficient of each tree clustering structure with different numbers of clusters according to the following formula:

[0030]

[0031] In the formula, S t-k is the silhouette coefficient of the clustering result when k clusters are delineated in the t-th tree clustering structure; x u In this clustering result, it belongs to cluster C u Any historical and cultural resource space entity; x q In this clustering result, it belongs to cluster C u Any historical and cultural resource space entity, x q ≠x u ;x p In this clustering result, it belongs to any other cluster C p Any historical and cultural resource space entity, where C p ≠C u ; is x u to x p The average of the Euclidean distances between is x u to x q The average of the Euclidean distances between

[0032] Step 4.3: Compare the changes in silhouette coefficient with the increase in the number of clusters in different tree-like clustering structures; select the number of clusters and its tree-like clustering structure corresponding to the maximum silhouette coefficient as the classification result of historical and cultural resources within the target range.

[0033] Furthermore, the step 5 specifically includes:

[0034] Step 5.1, apply the multidimensional scaling algorithm, calculate the three-dimensional feature vector of each historical and cultural resource space entity according to the "historical and cultural resource-historical and cultural resource" similarity distance matrix, and use the three-dimensional feature vector as the coordinate of a scattered point to draw a three-dimensional scatter plot corresponding to the historical and cultural resources within the target range;

[0035] Step 5.2: Perform natural language processing on the text of the data related to each historical and cultural resource space entity, respectively capture the protection importance level and cultural tourism popularity information of each historical and cultural resource space entity, and quantify them into quantitative data. Finally, retrieve the administrative division where each historical and cultural resource space entity is located, and number the administrative division according to the preset rules;

[0036] Step 5.3: On the three-dimensional scatter plot, different colors are assigned to the scatter points according to the administrative division number or classification results, and the size of the scatter points is controlled according to the protection importance level or cultural and tourism popularity, so as to complete the visualization of the classification results of the diverse historical and cultural resources.

[0037] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above method.

[0038] A computer-readable storage medium having a computer program / instruction stored thereon, wherein the computer program / instruction implements the above method when executed by a processor.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. This invention realizes the efficient classification of a large number of diverse historical and cultural resource space entities through intelligent natural language processing and data mining technology, transforming the traditional "subjective evaluation" into "objective analysis", significantly improving the accuracy, objectivity and efficiency of classification;

[0041] 2. This invention innovatively constructs a clustering method for the similarity comparison of historical and cultural resources, updates the working methodology, and makes up for the one-sided and inefficient problems of previous cultural assessment work; compared with traditional methods, the assessment work efficiency can be increased by about 45%;

[0042] 3. The present invention provides users with an intuitive analysis tool through the visualization technology of three-dimensional scatter plots; at the same time, the present invention can adjust parameters in real time according to needs, thereby achieving flexibility and applicability of classification.

[0043] 4. The technical solution of the present invention has broad application prospects. It is not only suitable for the classification of historical and cultural resources, but can also be extended to the classification of other types of cultural heritage. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of a classification method of multiple historical and cultural resources provided by an embodiment of the present invention;

[0045] Figure 2 is a schematic diagram of a clustering tree structure generated by the present invention;

[0046] Figure 3 This is a comparison diagram of the silhouette coefficients of different clustering tree structures generated by the present invention.

[0047] Figure 4 is a three-dimensional scatter plot generated by the present invention;

[0048] Figure 5 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0050] Embodiment 1

[0051] The embodiment of the present invention provides a classification method for multiple historical and cultural resources, such as Figure 1 As shown, the following steps are included:

[0052] Step 1: Obtain the spatial entities belonging to historical and cultural resources within the target range and generate a list of historical and cultural resource spatial entity entries.

[0053] In specific implementation, the steps include:

[0054] Step 1.1: Collect relevant texts related to the protection and inheritance of historical and cultural resources, and extract the spatial entity types belonging to historical and cultural resources through natural language processing.

[0055] Among them, the relevant texts involving the protection and inheritance of historical and cultural resources can be the relevant policy texts on the protection and inheritance of historical and cultural resources, such as the relevant laws and regulations on the protection of historical and cultural resources published on the official websites of cultural relics bureaus at all levels and in various places across the country in the past 10 years, the public notices of cultural relics protection units, and the policies for specific types of historical and cultural resources. Through the natural language processing of text mining of "building a special dictionary + machine learning segmentation + named entity recognition", the spatial entity types belonging to historical and cultural resources are identified, as shown in Table 1.

[0056] Table 1: Types of spatial entities belonging to historical and cultural resources

[0057] Types of historical and cultural resources Historical and cultural resource space entity types Regulations: Cultural relics protection units National key / provincial / municipal and county-level cultural relics protection units Area: Historical and cultural town Historical and cultural city / block / town / village Cluster: Related towns and villages Traditional Village Entity: Related Buildings Historic buildings / immovable cultural relics / national industrial heritage Elements: related spatial environment elements Ancient well / ancient bridge / ancient tree / ancient place name

[0058] Step 1.2: According to the spatial entity types belonging to historical and cultural resources, obtain the spatial entities belonging to the spatial entity types of historical and cultural resources within the target range, and generate a list of historical and cultural resource spatial entity items.

[0059] Among them, the spatial entities belonging to the historical and cultural resource spatial entity type within the target range can be obtained by crawling electronic maps or collecting on-site surveys. Points of interest (POI) within the target range should include the name of the historical and cultural resource and its latitude and longitude coordinate information, administrative division information, identify the spatial entities belonging to the historical and cultural resource spatial entity type, merge all the identified spatial entities to obtain a list of historical and cultural resource spatial entity items, and number each historical and cultural resource spatial entity, which can be H 1 , H 2 , H 3 … Taking Nanjing as an example, the list of historical and cultural resource space entity items is shown in Table 2.

[0060] Table 2: List of historical and cultural space physical resource items within the target area

[0061] serial number Name of historical and cultural resource space entity <![CDATA[H 1 ]]> Confucius Temple Qinhuai River Scenic Area <![CDATA[H 2 ]]> Former Residence of Fu Baoshi <![CDATA[H 3 ]]> Ganxi Residence <![CDATA[H 4 ]]> Former site of Central Broadcasting Station <![CDATA[H 5 ]]> Cangxi Stage <![CDATA[H 6 ]]> Zhanyuan <![CDATA[H 7 ]]> Jiangnan Imperial Examination Hall … …

[0062] Step 2: According to the list of historical and cultural resource space entity entries, retrieve the relevant data text of the historical and cultural resource space entity, and extract the cultural characteristic dimensions of the historical and cultural resources, the cultural characteristic dimension weights and the cultural attributes covered by each cultural characteristic dimension through natural language processing of the relevant data text, and construct a "historical and cultural resources-cultural attributes" relationship mapping table and a "cultural attributes-cultural attributes" similarity matrix under each cultural dimension.

[0063] In specific implementation, the steps include:

[0064] Step 2.1, obtain several data texts related to each historical and cultural resource space entity in the list of historical and cultural resource space entity entries in the existing data documents, and form a related text set with all data texts related to all historical and cultural resource space entities.

[0065] Among them, several data texts related to historical and cultural resource space entities can be obtained in the following way: enter the name of the historical and cultural resource space entity into common Internet search engines and authoritative academic platforms for retrieval, and take the top 100 search results in the comprehensive ranking as relevant data texts.

[0066] Step 2.2: Perform natural language processing on the relevant text collection to extract several cultural feature dimensions and corresponding weights related to historical and cultural resources, as well as all cultural attributes covered by each cultural feature dimension.

[0067] Among them, the relevant text collection can be processed through the natural language processing scheme of "text cleaning + LDA probability topic model training + topic probability statistics", based on indicators such as keyword capture, word frequency statistics, and word meaning association, to extract several cultural feature dimensions and corresponding weights related to historical and cultural resources. Among them, the weight W i The value should be a real number between (0,1), i = 1, 2...n, and ensure Then, through the natural language processing solution of "text classification + word meaning association + keyword capture", all cultural attributes covered by each cultural dimension are obtained. For example, continuing with the previous example, three cultural characteristic dimensions are extracted from the list of historical and cultural resource space entity items in Table 2, namely: characteristic object type, historical figure type, and narrative method. According to the word frequency calculation and topic probability statistics, the proportion weights of the three cultural characteristic dimensions are obtained and numbered W in turn. 1 , W 2 , W 3 , as shown in Table 3, and all cultural attributes covered by each cultural characteristic dimension are shown in Table 4.

[0068] Table 3: Dimensions and weights of cultural characteristics

[0069] Cultural characteristics dimension Weight Feature type 0.42 Historical figure type 0.21 Narrative style 0.37

[0070] Table 4: Cultural attributes covered by each cultural characteristic dimension

[0071]

[0072] Step 2.3: Perform natural language processing on several data texts related to each historical and cultural resource space entity, extract the cultural attributes of each historical and cultural resource space entity in each cultural characteristic dimension, and establish a "historical and cultural resource-cultural attribute" relationship mapping table. The "historical and cultural resource-cultural attribute" relationship mapping table covers the mapping relationship between each historical and cultural resource space entity and all its cultural attributes.

[0073] For example, for the list of historical and cultural resource space entities in Table 2, the natural language processing results show that H 7 (Jiangnan Gongyuan) has the highest semantic correlation with the cultural attribute of "man-made structure" in "characteristic object type", the highest semantic correlation with "scholars and officials" in "historical figure type", and the highest semantic correlation with "Confucian narrative" in "narrative method". Therefore, H 7 The relationship between "historical and cultural resources-cultural attributes" of (Jiangnan Imperial Examination Hall) is mapped as "Jiangnan Imperial Examination Hall-man-made structures-literati-Confucian narratives".

[0074] In addition, for historical and cultural resource space entities with multiple cultural attributes under the same cultural characteristic dimension, the mapping relationship between the historical and cultural resource space entity and the different cultural attributes of the same cultural characteristic dimension is added as a mapping relationship to the "historical and cultural resource-cultural attribute" relationship mapping table, and the priority of multiple mapping relationships of the same cultural characteristic dimension is noted according to the degree of relevance. Multiple mapping relationships belonging to the same historical and cultural resource space entity form multiple sub-items belonging to the historical and cultural resource space entity. For example, the natural language processing results show that H 6 (Zhanyuan) has the highest semantic correlation with "place where historical events occurred" under "characteristic object type", and has a high semantic correlation with "bureaucrats", "political figures", and "gentry" under "historical figure type"; and has the highest semantic correlation with "gentry cultural narrative" under "narrative type". Therefore, H 6 (Zhanyuan) should have 1*3*1=3 sub-items, numbered as H 6_1 , H 6_2 , H 6_3 , and determine the primary and secondary relationships based on the degree of word meaning relevance (correlation). The final "historical and cultural resources-cultural attributes" relationship mapping table for the historical and cultural space entity resource item list in Table 2 is shown in Table 5.

[0075] Table 5: “Historical and cultural resources-cultural attributes” mapping table

[0076] serial number Historical and cultural resource space entity Feature type Historical figure type Narrative style Primary and secondary relationship <![CDATA[H 1_1 ]]> Confucius Temple Qinhuai River Scenic Area Place where historical events took place merchant Confucian Narrative Second-rate <![CDATA[H 1_2 ]]> Confucius Temple Qinhuai River Scenic Area Place where historical events took place Scholar-officials Confucian Narrative host <![CDATA[H 1_3 ]]> Confucius Temple Qinhuai River Scenic Area Place where historical events took place the masses Confucian Narrative Second-rate … … … … … … <![CDATA[H 2_1 ]]> Fu Baoshi's Former Residence Former Residences of Celebrities Literary Masters Artistic folk narrative host <![CDATA[H 3_1 ]]> Ganxi Residence Man-made structures Scholar-officials Gentry Cultural Narrative host <![CDATA[H 4_1 ]]> Former site of Central Broadcasting Station Man-made structures the masses Modern National Narratives host <![CDATA[H 5_1 ]]> Cangxi Stage Place where historical events took place Literary Masters Artistic folk narrative host <![CDATA[H 6_1 ]]> Zhanyuan Place where historical events took place bureaucracy Gentry Cultural Narrative host <![CDATA[H 6_2 ]]> Zhanyuan Place where historical events took place Politicians Gentry Cultural Narrative Second-rate <![CDATA[H 6_3 ]]> Zhanyuan Place where historical events took place The well-dressed aristocracy Gentry Cultural Narrative Second-rate H7 Jiangnan Imperial Examination Hall Man-made structures Scholar-officials Confucian Narrative host … … … … … …

[0077] Step 2.4: Calculate the semantic similarity between any two cultural attributes under each cultural feature dimension to form a "cultural attribute-cultural attribute" similarity matrix under each cultural feature dimension.

[0078] Among them, semantic similarity can be directly calculated through natural language processing. For greater accuracy, the final semantic similarity can also be obtained by combining expert scores, that is: obtain the similarity scores of each two cultural attributes from more than 10 experts in the field (the higher the similarity, the greater the score); calculate the cosine similarity (Cosine Similarity) between the semantic similarity calculation result and the expert similarity score result. When the result is ≥0.8, it can be regarded as the similarity calculation result being generally recognized, and the result is processed into a positive integer between (0,10] through a standard score (Z-score), and the gradient corresponds to the similarity from irrelevant to completely relevant; when the calculation result is <0.8, the calculation corpus, algorithm, scoring experts, etc. should be adjusted and calculated again until a generally recognized similarity calculation result is obtained.

[0079] In this embodiment, on the one hand, the semantic similarity of the cultural attributes in Table 4 is calculated through the technical solution of "corpus vectorization-bag-of-words model construction-deep learning matching"; on the other hand, the expert scores are distributed and collected through the mobile data interface, and the scoring results are imported into the 2308 version Excel formula system for statistical processing; on the basis of calculating the cosine similarity of the two = 0.83 to ensure that the similarity calculation results of the two aspects are generally recognized, after standardization processing through the 2308 version Excel formula system, the similarity matrix of "cultural attribute-cultural attribute" under each cultural feature dimension is determined as shown in Table 6-Table 8;

[0080] Table 6: Similarity matrix of “cultural attribute-cultural attribute” under feature type

[0081] Material Dimension Place where historical events took place Natural scenery belt Man-made structures Representative residential buildings Former Residences of Celebrities Place where historical events took place 10 8 6 4 2 Natural scenery belt - 10 7 5 1 Man-made structures - - 10 9 6 Representative residential buildings - - - 10 7 Former Residences of Celebrities - - - - 10

[0082] Table 7: Similarity matrix of “cultural attribute-cultural attribute” under historical figure type

[0083] Character Dimension merchant Politicians Martyrs The well-dressed aristocracy bureaucracy the masses Scholar-officials Literary Masters merchant 10 9 8 3 2 2 2 1 Politicians - 10 9 8 7 4 3 2 Revolutionary Martyr - - 10 9 8 5 4 2 The well-dressed aristocracy - - - 10 7 4 3 2 bureaucracy - - - - 10 8 5 3 the masses - - - - - 10 6 2 Scholar-officials - - - - - - 10 8 Literary Masters - - - - - - - 10

[0084] Table 8: Similarity matrix of “cultural attribute-cultural attribute” under narrative mode

[0085]

[0086] Step 3: Based on the "historical and cultural resources-cultural attributes" relationship mapping table and the "cultural attributes-cultural attributes" similarity matrix, calculate the similarity distance between any two historical and cultural resource spatial entities and construct the "historical and cultural resources-historical and cultural resources" similarity distance matrix.

[0087] In specific implementation, the steps include:

[0088] Step 3.1: Based on the “historical and cultural resources-cultural attributes” relationship mapping table and the “cultural attributes-cultural attributes” similarity matrix, obtain the similarity between any two historical and cultural resource space entities in each cultural feature dimension, and form a similarity data set with all similarity data.

[0089] The steps for forming the similarity data set are as follows: importing the “historical and cultural resources-cultural attributes” relationship mapping table and the “cultural attributes-cultural attributes” similarity matrix into the 2308 version of the Excel formula system; in a computer system with the Python environment of the pandas and openpyx libraries installed, permuting and combining each mapping relationship in the “historical and cultural resources-cultural attributes” relationship mapping table, among which, the sub-items belonging to the same historical and cultural resource space entity (such as H x_1 With H x_2 ) does not participate in permutations and combinations, and only the sub-items with a primary and secondary relationship of "primary" need to be permuted. Through permutations and combinations, the numbered columns 1 and 2 shown in Table 8 are obtained; the vlookup function of the 2308 version of the Excel formula system is used to search for the cultural attributes mapped by each historical and cultural resource space entity and the similarity between each cultural attribute in Tables 6 to 8, and finally the L shown in Table 9 is obtained. 1 , L 2 , L 3 For example, from Table 5, we can see that H 1_1 and H 2_1 The cultural attributes mapped to “Feature Type” are “Places of Historical Events-Former Residences of Celebrities”. By using the vlookup function in Table 6, we can see that the similarity of “Places of Historical Events-Former Residences of Celebrities” is 2, that is, H 1_1 and H 2_1 The similarity of the feature type (L 1 ) is 2.

[0090] Table 9: Schematic diagram of the cultural attribute similarity dataset between historical and cultural resources (partial)

[0091]

[0092] Step 3.2: Based on the similarity dataset, measure the similarity distance between any two historical and cultural resource space entities according to the following formula:

[0093]

[0094] In the formula, d(H x ,H y ) is the historical and cultural resource space entity H x , H y The similarity distance, p is the total number of cultural feature dimensions, W i is the weight of the i-th cultural characteristic dimension, L i is the similarity between any two historical and cultural resource space entities in the similarity dataset under the i-th cultural feature dimension, σ i is the standard deviation of all data under the i-th cultural feature dimension of all historical and cultural resource space entities in the similarity dataset. In this example, σ 1 is 2.673495491, σ 2 is 3.529113589, σ 3 is 2.571447422.

[0095] Step 3.3: Based on the similarity distances between any two historical and cultural resource space entities, construct a "historical and cultural resource-historical and cultural resource" similarity distance matrix; among which, the similarity distance between the same historical and cultural resource space entities is 0.

[0096] In this example, the similarity distance matrix of “historical and cultural resources-historical and cultural resources” is shown in Table 10.

[0097] Table 10: Similarity distance matrix of “historical and cultural resources-historical and cultural resources” (partial)

[0098] <![CDATA[H 1_1 ]]> <![CDATA[H 2_1 ]]> <![CDATA[H 3_1 ]]> <![CDATA[H 4_1 ]]> … <![CDATA[H 1_1 ]]> 0 2.354098959 2.308691538 1.444989528 … <![CDATA[H 2_1 ]]> / 0 0.460136166 1.498929186 … <![CDATA[H 3_1 ]]> / / 0 1.444989528 … <![CDATA[H 4_1 ]]> / / / 0 … … … … … … …

[0099] Step 4: Based on the “historical and cultural resources-historical and cultural resources” similarity distance matrix, different clustering methods are used to generate different tree clustering structures, and the tree clustering structure with the largest silhouette coefficient and the number of clusters are selected as the classification result of historical and cultural resources.

[0100] In specific implementation, the steps include:

[0101] Step 4.1: Based on the “historical and cultural resources-historical and cultural resources” similarity distance matrix, clustering is performed using a plurality of clustering algorithms with different standards to obtain a plurality of clustering tree structures.

[0102] Among them, clustering algorithms with different standards include applying hierarchical clustering algorithms with complete link method, shortest distance method, average distance method, least square difference method, etc. as distance standards, obtaining different clustering results, and drawing cluster dendrograms under different clustering structures. For example, Figure 2 Shown is the cluster dendrogram when the distance criterion is the average distance method.

[0103] Step 4.2: Calculate the silhouette coefficient of each tree clustering structure with different numbers of clusters according to the following formula:

[0104]

[0105] In the formula, S t-k is the silhouette coefficient of the clustering result when k clusters are delineated in the t-th tree clustering structure; x u In this clustering result, it belongs to cluster C u Any historical and cultural resource space entity; x q In this clustering result, it belongs to cluster C u Any historical and cultural resource space entity, x q ≠x u ;x p In this clustering result, it belongs to any other cluster C p Any historical and cultural resource space entity, where C p ≠C u ; is x u to x p The average of the Euclidean distances between is x u to x q The average of the Euclidean distances between

[0106] Step 4.3: Compare the changes in silhouette coefficient with the increase in the number of clusters in different tree-like clustering structures; select the number of clusters and its tree-like clustering structure corresponding to the maximum silhouette coefficient as the classification result of historical and cultural resources within the target range.

[0107] In addition, abnormal / isolated individuals of historical and cultural resource spatial entities should be extracted based on the clustering dendrogram, and verified by comparison to complete the clustering of diverse historical and cultural resources.

[0108] In this embodiment, the clustering result and the calculation result can be obtained by substituting the R language into the function calculation, and the calculated clustering result silhouette coefficient is as follows: Figure 3 As shown in Table 11 and Figure 3The classification results shown;

[0109] Table 11: Classification results of historical and cultural resources (partial)

[0110] Historical and cultural resource space entity number Classification cluster number <![CDATA[H 1_1 ]]> 5 <![CDATA[H 1_2 ]]> 5 <![CDATA[H 1_3 ]]> 5 … … <![CDATA[H 2_1 ]]> 1 <![CDATA[H 3_1 ]]> 1 <![CDATA[H 4_1 ]]> 6 <![CDATA[H 5_1 ]]> 2 <![CDATA[H 6_1 ]]> 2 <![CDATA[H 6_2 ]]> 2 <![CDATA[H 6_3 ]]> 2 <![CDATA[H 7_1 ]]> 2 <![CDATA[H 8_1 ]]> 8 <![CDATA[H 8_2 ]]> 8 <![CDATA[H 9_1 ]]> 6 <![CDATA[H 9_2 ]]> 6 <![CDATA[H 9_3 ]]> 6 <![CDATA[H 10_1 ]]> 8 <![CDATA[H 10_2 ]]> 8 … …

[0111] Step 5: Based on the “historical and cultural resources-historical and cultural resources” similarity distance matrix, draw a three-dimensional scatter plot that can present the classification results, protection importance, administrative location, and cultural and tourism popularity of the historical and cultural resource spatial entities.

[0112] In specific implementation, the steps include:

[0113] Step 5.1: Apply the multidimensional scaling algorithm, calculate the three-dimensional feature vector of each historical and cultural resource spatial entity based on the "historical and cultural resources-historical and cultural resources" similarity distance matrix, and use the three-dimensional feature vector as the coordinate of a scatter point to draw a three-dimensional scatter graph (3-D scatter graph) corresponding to the historical and cultural resources within the target range.

[0114] Among them, the specific method for generating a three-dimensional scatter plot is as follows: in a computer system with an R language environment with the stats library installed, use the "cmdscale(,k=3)" function to calculate the similarity distance matrix as a three-dimensional feature vector of each historical and cultural resource spatial entity, as shown in Table 12; then import the feature vector into the Origin 2022 software system, set the three dimensions of the feature vector as X, Y, and Z respectively, convert the feature vector into a three-dimensional Cartesian coordinate, and use the "Drawing-3D Scatter Plot" function to draw a three-dimensional scatter plot of historical and cultural attributes.

[0115] Table 12: Three-dimensional feature vectors of each historical and cultural resource space entity (partial)

[0116]

[0117]

[0118] Step 5.2: Perform natural language processing on the text of the data related to each historical and cultural resource space entity, capture the protection importance level and cultural and tourism popularity information of each historical and cultural resource space entity, and quantify them into quantitative data. Finally, retrieve the administrative divisions where each historical and cultural resource space entity is located, and number the administrative divisions according to preset rules.

[0119] Among them, the protection importance level and cultural and tourism popularity information of each historical and cultural resource can be obtained through the natural language processing technology solution of "data cleaning-keyword matching and text crawling-word frequency statistics", as shown in the 2nd and 3rd columns from the left of Table 13; the administrative division numbering rules can be 1, 2, 3, or they can be numbered in the form of A, B, C, etc. In this example, the sequential numbering of 1, 2, 3 is adopted, as shown in the 4th row from the left of Table 13.

[0120] Table 13: Popularity data, importance level and administrative division code of historical and cultural resource spatial entities (partial)

[0121]

[0122]

[0123] Step 5.3: On the three-dimensional scatter plot, different colors are assigned to the scatter points according to the administrative division number or classification results, and the size of the scatter points is controlled according to the protection importance level or cultural and tourism popularity, so as to complete the visualization of the classification results of the diverse historical and cultural resources.

[0124] For example, the same color is assigned to the points with the same administrative division number or classification result. For example, the higher the importance level, the higher the cultural and tourism popularity, and the larger the scatter point. Figure 4 As shown. They are: 1) Cluster evaluation results control color, protection importance controls size; 2) Cluster evaluation results control color, cultural tourism heat controls size; 3) Administrative division number controls color, protection importance controls size; 4) Administrative division number controls color, cultural tourism heat controls size.

[0125] Three-dimensional scatter plots can provide decision-making assistance for the protection and planning of historical and cultural resources. Figure 3 It can be seen that the points with zoning numbers 8, 9, and 10 are clustered together, indicating that these historical and cultural resources are closer in terms of cultural characteristics, that is, they have more similar material meanings, character characteristics, and narrative backgrounds, and are therefore more suitable for inclusion in the same protection planning design. At the same time, in these points, the spatial entity size of historical and cultural resources is unevenly distributed, so design resources need to be allocated accordingly.

[0126] Embodiment 2

[0127] Figure 5 Schematic diagram of the structure of a computer device provided by an embodiment of the present invention. The embodiment of the present invention provides services for implementing the method of the first embodiment of the present invention. Figure 5As shown, the device may include: a memory 301 storing a computer executable program; a processor 302 coupled to the memory 301; the processor 302 calls the computer executable program stored in the memory 301 to execute the steps in the method described in the first embodiment.

[0128] The memory 301 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The device may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the memory 301 may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as "hard drives"). A program / utility having a set (at least one) of program modules may be stored in, for example, the memory 301, such program modules including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include the implementation of a network environment. The computer executable program of the program module typically performs the functions and / or methods in the embodiments described in the present invention.

[0129] The processor 302 executes various functional applications and data processing by running the program stored in the memory 301, such as implementing the method provided in the first embodiment of the present invention.

[0130] The code of the computer executable program can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" language or similar programming languages.

[0131] Embodiment 3

[0132] An embodiment of the present invention provides a storage medium including a computer executable program. When the computer executable program is executed by a computer processor, it is used to execute the method of the first embodiment.

[0133] The storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, a device or a device or used in combination with it.

[0134] Of course, the storage medium containing a computer executable program provided by an embodiment of the present invention, the computer executable program of which is not limited to the above method operations, can also execute related operations in the method provided by any embodiment of the present invention.

[0135] It should be understood that the above embodiments and descriptions only describe the principles, main features and advantages of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, and these changes and improvements all fall within the scope of protection of the present invention.

Claims

1. A classification method for multiple historical and cultural resources, characterized in that: The steps include: Step 1: Obtain the spatial entities belonging to historical and cultural resources within the target range and generate a list of historical and cultural resource spatial entity items; Step 2: According to the list of historical and cultural resource space entity entries, retrieve the relevant data text of the historical and cultural resource space entity, and extract the cultural characteristic dimensions of the historical and cultural resources, the cultural characteristic dimension weights and the cultural attributes covered by each cultural characteristic dimension through natural language processing of the relevant data text, and finally construct a "historical and cultural resources-cultural attributes" relationship mapping table and a "cultural attributes-cultural attributes" similarity matrix under each cultural characteristic dimension; Step 3: According to the "historical and cultural resources-cultural attributes" relationship mapping table and the "cultural attributes-cultural attributes" similarity matrix, the similarity distance between any two historical and cultural resource space entities is calculated, and the "historical and cultural resources-historical and cultural resources" similarity distance matrix is ​​constructed; Step 4: Based on the "historical and cultural resources-historical and cultural resources" similarity distance matrix, different clustering methods are used to generate different tree clustering structures, and the tree clustering structure with the largest silhouette coefficient and the number of clusters are selected as the classification result of historical and cultural resources; The step 2 specifically includes: Step 2.1, obtain several data texts related to each historical and cultural resource space entity in the historical and cultural resource space entity entry list from existing data documents, and form a related text set with all data texts related to all historical and cultural resource space entities; Step 2.2, perform natural language processing on the relevant text set, extract several cultural feature dimensions and corresponding weights related to historical and cultural resources through the natural language processing solution of "text cleaning + LDA probabilistic topic model training + topic probability statistics", and obtain all cultural attributes covered by each cultural feature dimension through the natural language processing solution of "text classification + word meaning association + keyword crawling"; Step 2.3: Perform natural language processing on a number of data texts related to each historical and cultural resource space entity, extract the cultural attributes of each historical and cultural resource space entity in each cultural feature dimension, and establish a "historical and cultural resource-cultural attribute" relationship mapping table, which covers the mapping relationship between each historical and cultural resource space entity and all its cultural attributes; Step 2.4: Calculate the semantic similarity between any two cultural attributes under each cultural feature dimension to form a "cultural attribute-cultural attribute" similarity matrix under each cultural feature dimension.

2. The classification method of multiple historical and cultural resources according to claim 1 is characterized in that: The method further comprises the steps of: Step 5. Based on the "historical and cultural resources-historical and cultural resources" similarity distance matrix, draw a three-dimensional scatter plot that can present the classification results, protection importance, administrative location, and cultural and tourism popularity of historical and cultural resource spatial entities.

3. The classification method of multiple historical and cultural resources according to claim 1 is characterized in that: The step 1 specifically includes: Step 1.1: Collect relevant texts related to the protection and inheritance of historical and cultural resources, and extract the spatial entity types belonging to historical and cultural resources through natural language processing; Step 1.2: According to the spatial entity types belonging to historical and cultural resources, obtain the spatial entities belonging to the spatial entity types of historical and cultural resources within the target range, and generate a list of historical and cultural resource spatial entity items.

4. The classification method of multiple historical and cultural resources according to claim 1 is characterized in that: In step 2.3, for the historical and cultural resource space entity with multiple cultural attributes under the same cultural characteristic dimension, the mapping relationship between the historical and cultural resource space entity and the different cultural attributes of the same cultural characteristic dimension is added as a mapping relationship to the "historical and cultural resources-cultural attributes" relationship mapping table, and the priority of multiple mapping relationships in the same cultural characteristic dimension is noted according to the degree of relevance.

5. The classification method of multiple historical and cultural resources according to claim 1 is characterized in that: The step 3 specifically includes: Step 3.1, according to the "historical and cultural resources-cultural attributes" relationship mapping table and the "cultural attributes-cultural attributes" similarity matrix, obtain the similarity of any two historical and cultural resource space entities in each cultural feature dimension, and form all similarity data into a similarity data set; Step 3.2: Based on the similarity dataset, measure the similarity distance between any two historical and cultural resource space entities according to the following formula: In the formula, d(H x ,H y ) is the historical and cultural resource space entity H x , H y The similarity distance, p is the total number of cultural feature dimensions, W i is the weight of the i-th cultural characteristic dimension, L i is the similarity between any two historical and cultural resource space entities in the similarity dataset under the i-th cultural feature dimension, σ i is the standard deviation of all data under the i-th cultural feature dimension of all historical and cultural resource space entities in the similarity dataset; Step 3.3: Based on the similarity distances between any two historical and cultural resource space entities, construct a "historical and cultural resource-historical and cultural resource" similarity distance matrix; among which, the similarity distance between the same historical and cultural resource space entities is 0.

6. The classification method of multiple historical and cultural resources according to claim 1 is characterized in that: The step 4 specifically includes: Step 4.1: Based on the "historical and cultural resources-historical and cultural resources" similarity distance matrix, clustering is performed using a plurality of clustering algorithms with different standards to obtain a plurality of clustering tree structures; Step 4.2: Calculate the silhouette coefficient of each tree clustering structure with different numbers of clusters according to the following formula: In the formula, S t-k is the silhouette coefficient of the clustering result when k clusters are delineated in the t-th tree clustering structure; x u In this clustering result, it belongs to cluster C u Any historical and cultural resource space entity; x q In this clustering result, it belongs to cluster C u Any historical and cultural resource space entity, where x q ≠x u ;x p In this clustering result, it belongs to any other cluster C p Any historical and cultural resource space entity, where C p ≠C u ; is x u to x p The average of the Euclidean distances between is x u to x q The average of the Euclidean distances between Step 4.3: Compare the changes in silhouette coefficient with the increase in the number of clusters in different tree-like clustering structures; select the number of clusters and its tree-like clustering structure corresponding to the maximum silhouette coefficient as the classification result of historical and cultural resources within the target range.

7. The classification method of multiple historical and cultural resources according to claim 2 is characterized in that: The step 5 specifically includes: Step 5.1, apply the multidimensional scaling algorithm, calculate the three-dimensional feature vector of each historical and cultural resource space entity according to the "historical and cultural resource-historical and cultural resource" similarity distance matrix, and use the three-dimensional feature vector as the coordinate of a scattered point to draw a three-dimensional scatter plot corresponding to the historical and cultural resources within the target range; Step 5.2: Perform natural language processing on the text of the relevant data of each historical and cultural resource space entity, respectively capture the protection importance level and cultural tourism popularity information of each historical and cultural resource space entity, and quantify them into quantitative data. Finally, retrieve the administrative division where each historical and cultural resource space entity is located, and number the administrative division according to the preset rules; Step 5.3: On the three-dimensional scatter plot, different colors are assigned to the scatter points according to the administrative division number or classification results, and the size of the scatter points is controlled according to the protection importance level or cultural and tourism popularity, so as to complete the visualization of the classification results of the diverse historical and cultural resources.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the computer program to implement the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: The computer program / instructions, when executed by a processor, implement the method of any one of claims 1-7.

Images