Knowledge graph-based grotto image generation interruption analysis method

Abstract

The invention discloses a grotto image generation interruption analysis method based on a knowledge graph, and belongs to the technical field of cultural heritage protection and analysis. The method comprises the following steps: S1, collecting feature information of grotto images through historical literatures, archaeological reports and digital resources, and processing the obtained data to construct a real physical world map; s2, further analyzing feature information of the grotto image, constructing a multi-dimensional relationship among features, and further constructing a conceptual map of the grotto image; s3, integrating element information related to grotto imaging, constructing a dynamic relationship among the element information, and comprehensively analyzing the element information to construct a grotto imaging affair map; and S4, integrating the real physical world map, the concept map and the fact map obtained in the steps S1 to S3 to construct a corresponding breakout analysis model, and completing breakout analysis work of grotto imaging. According to the method, a brand-new research normal form is provided for historical analysis and deduction under a complex cultural background by constructing the multi-level knowledge graph with the dynamic reasoning capability.

Description

Technical Field

[0001] The present invention relates to the technical field of cultural heritage protection and analysis, and specifically to a method for dating and analyzing cave statues based on a knowledge graph. Background Art

[0002] The traditional dating method of cave statues mainly relies on the experience of archaeological experts, who make judgments by observing the volume, shape, carving craftsmanship, clothing style and other characteristics of the statues, combined with historical documents. This method is inefficient, the results are easily affected by subjective factors, and it is difficult to process large-scale statue data. With the advancement of cultural heritage digital technology, there is an urgent need for an intelligent and systematic method that can start from the real data of the physical world and realize the automation and efficiency of cave statue dating through the construction and reasoning of multi-level knowledge graphs.

[0003] Contemporary digital cultural heritage research should not only be based on the research foundation of traditional archaeology and cultural heritage, but also needs to be combined with the application of advanced digital technology in this field. Therefore, the research object of digital cultural heritage is inseparable from the essence of cultural heritage research. The research object is the cultural relics left by ancient humans in social activities, and digital technology is the main tool in research technology and equipment. For example, scholars represented by Huggett Jeremy not only proposed that digital cultural heritage is based on the assistance of ever-changing equipment, technology, methodology, software, and hardware, but also reshaped the essence of digital cultural heritage research in their own way to varying degrees. However, it also points out the current challenges in this field. For example, digital cultural heritage and archaeological tools based on computer technology are not innovative. They are just digital tools and can also be used in other fields, and these digital tools have no substantial progress in improving the essence of archaeological disciplines. In view of the above reasons, the present invention proposes a method for dating and analyzing cave statues based on knowledge graphs. Summary of the invention

[0004] 1. Technical Problems to be Solved by the Present Invention

[0005] The purpose of the present invention is to provide a method for dating and analyzing cave statues based on knowledge graphs to solve the problems raised in the above-mentioned background technology. The present invention constructs and generates three types of knowledge graphs (real physical world graph, cave statue concept graph, and cave statue principle graph) to establish a systematic process from data to dating, thereby solving the problems of low efficiency and insufficient precision in dating analysis of cultural heritage and realizing the automation and explainability of dating results.

[0006] 2. Technical solution

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for dating cave statues based on knowledge graphs, comprising the following steps:

[0009] S1. Constructing a map of the real physical world: Collecting characteristic information of cave statues through historical documents, archaeological reports and digital resources, and processing the obtained data to construct a map of the real physical world;

[0010] S2. Generate a conceptual map of cave statues: Based on the real physical world map constructed in S1, further analyze the characteristic information of cave statues, construct a multi-dimensional relationship between the characteristics, and further construct a conceptual map of cave statues;

[0011] S3. Generate a causal map of cave statues: Integrate the element information related to cave statues, build a dynamic relationship between the element information, and conduct a comprehensive analysis of the element information to build a causal map of cave statues;

[0012] S4, period analysis: Integrate the real physical world map, concept map and principle map obtained in S1~S3 to construct a corresponding period analysis model to complete the period analysis of the cave statues.

[0013] Preferably, the S1 specifically includes the following contents:

[0014] S1.1, Data collection: Collect image data and physical features of cave statues, integrate historical documents, archaeological reports, and extract time, location, and material information;

[0015] S1.2, Graph construction: Create the core nodes of the graph and the relationships between related elements;

[0016] S1.3. Brief description of the association structure: Through the multi-dimensional associations between core nodes, the network relationship of the history, space, artistic style and cultural background of the cave statues is constructed;

[0017] Preferably, the core nodes in S1.2 include dynasty, location, inscription, height of the statue, face, clothing, hairstyle, pedestal, and backlight; the related element relationships include historical period and artistic style, artistic style and craftsmanship, craftsmanship and material, cultural influence and artistic style, cave location and historical period, religious belief and artistic style, distribution and regional characteristics.

[0018] Preferably, S2 specifically includes the following contents:

[0019] S2.1, Conceptual abstraction: Based on the entities and relationships in the real physical world map constructed in S1, extract high-level concepts, combine physical characteristics with cultural characteristics, generate a conceptual hierarchy with academic explanatory power, integrate the generated conceptual hierarchy, and enhance the theoretical depth of the concept map;

[0020] S2.2, Association mechanism construction: Establish multi-dimensional relationships in the concept map, and reveal the deep relationship between the style characteristics, cultural spirit and historical value of the cave statues from a multi-dimensional perspective;

[0021] S2.3, Image and text matching: Extract features from the images of cave statues and match them with relevant nodes in the concept map, align the images with descriptive texts, and match the images with multiple conceptual features of different regions and dynasties in the context of cultural integration to enhance the breadth and depth of analysis;

[0022] S2.4. Integration of archaeological aesthetics and art: Combining archaeological aesthetics with art theory, the artistic elements of cave statues are incorporated into the conceptual map. From the perspective of archaeological aesthetics, the correspondence between cave statues and the aesthetic standards of the era to which they belong is analyzed, and cultural background explanations are provided.

[0023] Preferably, the establishing of multidimensional relationships in the concept map in S2.2 specifically refers to establishing multidimensional relationships among history, space, time, people and folk culture, wherein:

[0024] History: Combine time, events, and dynasty changes to reveal the evolution of statue styles;

[0025] Space: Associate different regions and geographical relationships to analyze the distribution characteristics of cave statues;

[0026] Time: Through the timeline positioning, the correlation between the image style and time is deduced;

[0027] Characters: Combine statues, historical figures and their story backgrounds to enrich the content of the concept map;

[0028] Folk culture: Integrate folk cultural characteristics and explain the interactive relationship between cave statues and folk culture.

[0029] Preferably, S3 specifically includes the following contents:

[0030] S3.1. Dynamic relationship construction: integrating events, geographical space, and cultural customs to form a map of events and dynamically simulate the formation and dissemination process of cave statues;

[0031] S3.2. Comprehensive factor analysis: Integrate multi-dimensional information on characters, stories, space, and time to form a period model under a complex cultural background, and conduct in-depth exploration and analysis of regional characteristics, artistic style, folk culture, and historical value.

[0032] Preferably, the S4 specifically includes the following contents:

[0033] S4.1. Graph integration: Integrate the real physical world graph, concept graph and event graph to establish a multi-level association model and screen out key generation features;

[0034] S4.2, Automated dating: Input the cave statue images into the constructed multi-level association model to extract physical features; map the features to the knowledge graph to generate dating results; the output dating results include time, region, style features and interpretability analysis;

[0035] S4.3. Use the knowledge graph to drive the dating process: Through the precise matching of statue images with nodes in the knowledge graph, multi-dimensional reasoning is performed in combination with time, space, and cultural characteristics to provide logical support for dating and cultural background explanation;

[0036] S4.4. Cross-dimensional correlation: Establish cross-dimensional correlation between physical world data and conceptual and rational layers. Focusing on the phenomenon of cultural integration, combine the artistic features and aesthetic styles of Buddhist statues from different dynasties to explore the deep correlation between time and space.

[0037] The present invention further protects a knowledge graph-based cave statue dating analysis system for executing the above method.

[0038] The present invention further protects a computer device, which includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the instruction, program, code set or instruction set is loaded and executed by the processor to implement the above-mentioned knowledge graph-based cave statue dating analysis method.

[0039] 3. Beneficial effects

[0040] (1) The present invention proposes a method for constructing a multi-level knowledge graph (real physical world graph, concept graph, and fact graph), breaking through the limitations of traditional single graph analysis.

[0041] (2) The present invention integrates high-level concepts such as cultural spirit, value concepts, and art (archaeological aesthetics), which can enhance the cultural theoretical depth of period analysis.

[0042] (3) The present invention emphasizes the knowledge graph-driven dating process, which significantly improves the systematization and automation level of cultural heritage analysis.

[0043] (4) The present invention can improve the credibility of dating results by matching pictures with descriptive texts and combining image analysis with historical document verification.

[0044] (5) The present invention simplifies the complex periodization logic and refines the core process into an interpretable knowledge graph analysis framework.

[0045] In summary, the present invention can be widely used in the fields of cultural heritage protection, archaeological research, and museum digital display. Intelligent period analysis not only improves the efficiency of archaeological work, but also provides a scientific basis for the value assessment and protection planning of cultural heritage. In addition, this method is highly scalable and can be adapted to the study of other types of cultural heritage, such as architectural sites, historical documents, etc. The present invention provides a new research paradigm for historical analysis and periodization in a complex cultural background by constructing a multi-level knowledge graph with dynamic reasoning capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 This is a flow chart of a method for dating cave statues based on knowledge graph proposed by the present invention;

[0048] Figure 2 This is a schematic diagram of the physical map construction process mentioned in Example 1 of the present invention;

[0049] Figure 3 This is a schematic diagram of the process of constructing the concept map mentioned in Example 1 of the present invention;

[0050] Figure 4 A schematic diagram of the process of constructing the event graph mentioned in Example 1 of the present invention;

[0051] Figure 5 This is a schematic diagram of the conclusion of the dating analysis of the cave statues mentioned in Example 2 of the present invention. DETAILED DESCRIPTION

[0052] 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.

[0053] Embodiment 1:

[0054] See also Figure 1 The present invention proposes a method for dating cave statues based on knowledge graph, which specifically includes the following contents:

[0055] 1. Construction of a real physical world map

[0056] 1.1 Data source:

[0057] (1) Collect image data and physical characteristics of cave statues (such as volume, shape, carving technology, etc.);

[0058] (2) Integrate historical documents and archaeological reports to extract key information such as time, location, and materials.

[0059] 1.2 Graph construction:

[0060] See also Figure 2 Based on the collected data information, the obtained data is cleaned, different types of entities such as age and cave statues and the relationship between entities are clarified, and a map of the real physical world is constructed, which specifically includes the following contents:

[0061] (1) Create core nodes: dynasty, location, inscription, height, face, clothing, hairstyle, pedestal, backlight, etc.; taking Manjusri Bodhisattva as an example, the specific content is shown in Table 1:

[0062] Table 1

[0063]

[0064] (2) Relationships between related elements: historical period and artistic style, artistic style and craftsmanship, craftsmanship and materials, cultural influence and artistic style, cave location and historical period, religious beliefs and artistic style, distribution and regional characteristics, etc.

[0065] 1.3 Brief description of association structure:

[0066] Through the multi-dimensional associations between nodes, the network relationship of the history, space, artistic style and cultural background of the cave statues is constructed. The core elements are interrelated and verifiable, which improves the expressiveness and reasoning ability of the map.

[0067] 2. Generation of the Conceptual Atlas of Grotto Statues

[0068] See also Figure 3 For the construction of concept graph, the present invention is based on the constraint condition of "using the embedded vector representation corresponding to the description of the concept entity to support the relationship between the multimodal vector representation of the physical entity and the vector representation of the concept entity", and continuously optimizes the expression content of the concept entity (such as the concept Buddha statue), thereby constructing a set of association relationships with invisible relationships between the concept entity (concept Buddha statue) and the physical entity (era) ( Figure 3 The dotted line represents the invisible relationship, and the thickness of the dotted line represents the strength of the relationship).

[0069] Specifically, the present invention is based on a multimodal vector representation model, and obtains the embedded vector representations corresponding to the description texts of different concept entities (conceptual Buddha statues), as well as the embedded vector representations represented by the description information of different eras, and then measures the association relationship and relationship strength q between the two entities by calculating the "distance" (such as cosine similarity) between the vectors. The above process can be represented by the following formula:

[0070] vec 概念表征 =[a 1 ,a 2 ,…,a n ]

[0071] vec 年代表征 =[b 1 ,b 2 ,…,b n ]

[0072] Then the magnitude of the relationship strength q between two entities is:

[0073]

[0074] In the above formula, n represents the dimension of the embedding vector model;

[0075] vec 概念表征 ·vec 年代表征 It represents the inner product operation of two vectors; ||vec 概念表征 ||·||vec 年代表征 || represents the Euclidean norm (modulus) of two vectors.

[0076] Based on the above content, the concept map generation specifically includes the following contents:

[0077] 2.1 Conceptual Abstraction:

[0078] (1) Extract high-level concepts (such as Buddhist sect style, regional style, carving technology, etc.) based on the entities and relationships in the physical map;

[0079] (2) combining physical features with cultural characteristics to generate a conceptual hierarchy with academic explanatory power;

[0080] (3) Integrate advanced cultural dimensions such as cultural spirit, value concepts, and art (archaeological aesthetics) to enhance the theoretical depth of the concept map. Taking Manjusri Bodhisattva as an example, the specific content is shown in Table 2:

[0081] Table 2

[0082]

[0083] 2.2 Association mechanism:

[0084] (1) Establish multidimensional relationships in the concept map, including:

[0085] History: Combine time, events, and dynasty changes to reveal the evolution of statue styles;

[0086] Space: Associate different regions and geographical relationships to analyze the distribution characteristics of cave statues;

[0087] Time: Through the timeline positioning, the correlation between the image style and time is deduced;

[0088] Characters: Combine statues, historical figures and their story backgrounds to enrich the content of the concept map;

[0089] Folk culture: Integrate folk cultural characteristics and explain the interactive relationship between cave statues and folk culture;

[0090] (2) Reveal the deep connection between the style characteristics, cultural spirit and historical value of cave statues through a multi-dimensional perspective.

[0091] 2.3 Matching of images and text:

[0092] (1) Extract features from cave statue images and match them with relevant nodes in the concept graph;

[0093] (2) Support the alignment of images and descriptive text, for example, by analyzing the correspondence between the specific artistic style of Buddhist statues and historical records to verify the logic of dating;

[0094] (3) In the context of cultural integration, images are matched with the multiple conceptual features of different regions and dynasties to enhance the breadth and depth of analysis.

[0095] 2.4 Integrating archaeological aesthetics and art:

[0096] (1) Combining archaeological aesthetics with art theory, the artistic elements of cave statues, such as carving techniques, body structure, and aesthetic style, are incorporated into the conceptual map;

[0097] (2) From the perspective of archaeological aesthetics, the correspondence between cave statues and the aesthetic standards of the era to which they belonged is analyzed, and an explanation of the cultural background is provided.

[0098] 3. The Generation of the Schema of Grotto Statues

[0099] See also Figure 4For the construction of the event graph, further, through the above scheme, the present invention continuously optimizes the expression of the concept entity (such as the descriptive text information of the concept Buddha statue), firstly forms the concept entity representation based on the initial concept entity expression text, then calculates the relationship strength with the real-world age entity, obtains the relationship strength q between each concept entity and each age entity, then records the relationship strength data, manually verifies the relationship strength value, finds out the invalid relationship, then finds the corresponding description information of the corresponding concept entity according to the relationship, and modifies the description information of the concept entity.

[0100] Repeat the above process, generate the embedding vectors of concept entities and age entities again, recalculate the relationship strength between concept entities and real-world age entities, and then re-verify, continuously optimize the description text of concept entities, and finally form a more reasonable association relationship based on the above-mentioned "fact analysis" optimization process, that is, the dotted line association in the concept map is transformed into the solid line association relationship in the fact map, thereby constructing the corresponding fact map.

[0101] Of course, the above process can also be used to optimize the description information of the age entity in the real-world physical map. Based on the above content, the generation of the event map specifically includes the following contents:

[0102] 3.1 Dynamic Relationship:

[0103] (1) Integrate events (such as the change of dynasties and the spread of religion), geographical space, cultural customs and other elements to form a map of events;

[0104] (2) Dynamically simulate the formation and dissemination process of cave statues.

[0105] 3.2 Comprehensive factor analysis:

[0106] (1) Integrate multi-dimensional information such as characters, stories, space, and time to form a generation model in a complex cultural context;

[0107] (2) Includes in-depth exploration and analysis of factors such as regional characteristics, artistic style, folk culture, and historical value; the specific contents are shown in Table 3.

[0108] Table 3

[0109]

[0110]

[0111] 4. From Spectrum to Chronological Analysis

[0112] The statues are analyzed for dating using the real physical world map, concept map and principle map.

[0113] For the dating method of the real physical world map, the present invention uses a multimodal vector representation model to obtain the vector representation vec of the user input picture. input_img Then, the cosine similarity distance is calculated with all the age representation vector representations in the real physical world map. Using the K-nearest neighbor algorithm (KNN), the age characteristics corresponding to the input image information can be obtained, and the dating analysis can be realized.

[0114] Specifically, in the above process:

[0115] vec input_img =[v 1 ,v 2 ,...,v n ]

[0116] vec 年代表征i =[b i1 ,b i2 ,...,b in ]

[0117]

[0118] In the above formula, q i This represents the strength of the relationship between the input image and the i-th era in the real physical world map. Then, with the help of the K-nearest neighbor algorithm (KNN), the era to which the input image belongs can be obtained, thereby achieving era analysis.

[0119] For the generation method of concept graph, we use the multimodal vector representation model to obtain the vector representation vec of the user input image input_img , and then calculate the cosine similarity distance with all the concept entity vector representations in the concept graph. Using the K-nearest neighbor algorithm (KNN), the concept entity corresponding to the input image information can be obtained. Then, combined with the relationship between the concept entity and the age entity in the concept graph, the period analysis can be realized.

[0120] Specifically, in the above process:

[0121] vec input_img =[v 1 ,v 2 ,...,v n ]

[0122] vec 概念表征i =[a i1 ,a i2 ,...,a in ]

[0123]

[0124] In the above formula, q iIt represents the strength of the relationship between the input image and the i-th concept entity in the concept graph. Then, with the help of the KNN algorithm, the concept Buddha statue that best matches the input image can be obtained. Then, with the help of the relationship between the concept entity and the age entity in the concept graph, the dating analysis can be realized (Note: dating based on concept entities also has the characteristic of being invisible, because the relationship in the concept graph has the characteristic of being invisible).

[0125] For the dating method of the event graph, similarly, the present invention uses a multimodal vector representation model to obtain the vector representation vec of the user input picture input_img , and then calculate the cosine similarity distance with all the concept entity vector representations in the causal graph. Using the K-nearest neighbor algorithm (KNN), the concept entity corresponding to the input image information can be obtained. Then, combined with the relationship between the concept entity and the age entity in the causal graph, the period analysis can be realized.

[0126] Specifically, in the above process:

[0127] vec input_img =[v 1 ,v 2 ,...,v n ]

[0128] vec 概念表征i =[a i1 ,a i2 ,...,a in ]

[0129]

[0130] In the above formula, q i It represents the strength of the relationship between the input image and the i-th concept entity in the event graph. Then, with the help of the KNN algorithm, the most matching concept Buddha statue of the input image can be obtained. Then, with the help of the relationship between the concept entity and the age entity in the event graph formed by event analysis optimization, the generation analysis can be realized.

[0131] Based on the above content, the period analysis specifically includes the following contents

[0132] 4.1 Graph Integration:

[0133] (1) Integrate the real physical world map, concept map and event map to establish a multi-level association model;

[0134] (2) Screen out key dating features.

[0135] 4.2 Automated generation:

[0136] (1) Input cave statue images and extract physical features;

[0137] (2) Mapping features to the knowledge graph to generate generation results;

[0138] (3) The output results include time, region, style characteristics and interpretability analysis.

[0139] 4.3 Knowledge graph-driven generation process:

[0140] (1) Through the precise matching of imagery and nodes in the knowledge graph, multi-dimensional reasoning is performed by combining time, space, and cultural characteristics;

[0141] (2) Provide logical support and cultural background explanation for the periodization.

[0142] 4.4 Cross-dimensional association:

[0143] (1) Establish cross-dimensional associations between physical world data and the conceptual and logical layers;

[0144] (2) In view of the phenomenon of cultural integration, this paper combines the artistic characteristics and aesthetic styles of Buddhist statues from different dynasties to explore the deep connection between time and space.

[0145] Embodiment 2:

[0146] Based on Example 1, but different in that, the present invention further illustrates the cave statue dating analysis method based on knowledge graph proposed by the present invention with specific examples, and the specific contents are as follows.

[0147] 1. Extract key features

[0148] Appearance features:

[0149] Face: The face is fleshy, round and full, which conforms to the characteristics of "round face and high eyebrows" in the Yuan Dynasty.

[0150] Hairstyle: A high, fleshy bun with spiral hair, combined with a crown decoration.

[0151] Hand gestures and ritual implements: The left hand is on the chest and the right hand holds a sword. This kind of ritual implement and gesture is more common in the statues of Manjushri Bodhisattva from the Yuan Dynasty.

[0152] Clothing style: The decoration is delicate and smooth yet solemn, reflecting the style of "complex lines and gorgeous clothes" in Yuan Dynasty statues.

[0153] Backlight and environment: The backlight is delicately carved, and the upturned lotus pedestal is decorated with beads, expressing the dignity and sacredness of the Bodhisattva statue.

[0154] 2. Matching knowledge graph:

[0155] (1) Map of the real physical world

[0156] Yuan Dynasty nodes: associated with the Yuan Dynasty style in the graph, verified by the following features:

[0157] Facial features: round and full facial shape.

[0158] Instruments / gestures: The sword and wisdom sutra are the typical objects held by Manjushri Bodhisattva.

[0159] Clothing features: The lines of the clothes are delicate, gorgeous and luxurious, reflecting the aesthetic characteristics of Yuan Dynasty statues.

[0160] (2) Concept Map

[0161] Match the physical features extracted above with the concept graph:

[0162] Historical period and artistic style: Yuan Dynasty statues combined the artistic styles of Han Buddhism and Tibetan Buddhism, characterized by solemn shapes and gorgeous costumes.

[0163] Religious belief node: The image of Manjusri Bodhisattva emphasizes the symbol of wisdom and is commonly seen in Yuan Dynasty statues.

[0164] The carving style of Buddhist statues in the Yuan Dynasty focuses on three-dimensionality and layering, which is consistent with the overall modeling characteristics of this statue.

[0165] (3) Schema of Events

[0166] Statue location and historical period: The statue is located on Feilai Peak, which was still an important statue-making site during the Yuan Dynasty, reflecting the inheritance of Buddhist art during this period.

[0167] Cultural influence: During the Yuan Dynasty, the style of Buddhist sculpture was influenced by Tibetan Buddhism. Manjushri Bodhisattva, as a symbol of wisdom, gradually became an important subject, and ritual instruments and gestures became more prominent.

[0168] 3. Conclusion of dating:

[0169] Based on the above analysis and the key information derived from the knowledge graph, it is concluded that the statue is a statue of Manjusri Bodhisattva from the Yuan Dynasty. The main evidence includes (such as Figure 5 shown):

[0170] 1. Facial features: round and full, with high eyebrows and eyes, in line with the style of the Yuan Dynasty.

[0171] 2. Hand gestures and ritual implements: Holding the sword of wisdom and scriptures is a symbol of Manjushri Bodhisattva.

[0172] 3. Clothing style: The lines are delicate and complex, gorgeous and rich in layers, reflecting the aesthetic characteristics of the Yuan Dynasty.

[0173] 4. Historical background: As a Buddhist holy place, Feilai Peak continued to be important in the statue culture of the Yuan Dynasty, and its style was influenced by Tibetan Buddhism.

[0174] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and improved concepts of the present invention within the technical scope disclosed by the present invention, and they should be covered by the protection scope of the present invention.

Claims

1. A method for dating cave statues based on knowledge graph, characterized in that: The following steps are involved: S1. Constructing a map of the real physical world: Collecting characteristic information of cave statues through historical documents, archaeological reports and digital resources, and processing the obtained data to construct a map of the real physical world; S2. Generate a conceptual map of cave statues: Based on the real physical world map constructed in S1, further analyze the characteristic information of cave statues, construct a multi-dimensional relationship between the characteristics, and further construct a conceptual map of cave statues; S3. Generate a causal map of cave statues: Integrate the element information related to cave statues, build a dynamic relationship between the element information, and conduct a comprehensive analysis of the element information to build a causal map of cave statues; S4, period analysis: Integrate the real physical world map, concept map and principle map obtained in S1 to S3 to construct a corresponding period analysis model to complete the period analysis of the cave statues.

2. According to the method for dating cave statues based on knowledge graph according to claim 1, it is characterized in that: The S1 specifically includes the following contents: S1.1, Data collection: Collect image data and physical features of cave statues, integrate historical documents, archaeological reports, and extract time, location, and material information; S1.2, Graph construction: Create the core nodes of the graph and the relationships between related elements; S1.

3. Brief description of the association structure: Through the multi-dimensional associations between core nodes, a network relationship of the history, space, artistic style and cultural background of the cave statues is constructed.

3. The method for dating cave statues based on knowledge graph according to claim 2 is characterized in that: The core nodes described in S1.2 include dynasty, location, inscription, height of the statue, face, clothing, hairstyle, pedestal, and backlight; the related element relationships include historical period and artistic style, artistic style and craftsmanship, craftsmanship and material, cultural influence and artistic style, cave location and historical period, religious beliefs and artistic style, distribution and regional characteristics.

4. The method for dating cave statues based on knowledge graph according to claim 3 is characterized in that: The S2 specifically includes the following contents: S2.1, Conceptual abstraction: Based on the entities and relationships in the real physical world map constructed in S1, extract high-level concepts, combine physical characteristics with cultural characteristics, generate conceptual entities with academic explanatory power, integrate the generated conceptual entities, and enhance the theoretical depth of the concept map; S2.2, Association mechanism construction: Establish multi-dimensional relationships in the concept map, and reveal the deep connection between the style characteristics, cultural spirit and historical value of the cave statues from a multi-dimensional perspective; S2.3, Image and text matching: Extract features from the images of cave statues and match them with relevant nodes in the concept map, align the images with descriptive texts, and match the images with multiple conceptual features of different regions and dynasties in the context of cultural integration to enhance the breadth and depth of analysis; S2.

4. Integration of archaeological aesthetics and art: Combining archaeological aesthetics with art theory, the artistic elements of cave statues are incorporated into the conceptual map. From the perspective of archaeological aesthetics, the correspondence between cave statues and the aesthetic standards of the era to which they belong is analyzed, and cultural background explanations are provided.

5. The method for dating cave statues based on knowledge graph according to claim 4 is characterized in that: The S2 further includes the following contents: Based on a multimodal vector representation model, we obtain the embedded vector representations corresponding to the description texts of different conceptual entities, as well as the embedded vector representations represented by the description texts of different eras. We calculate the "distance" between the vectors to measure the association and relationship strength between the two entities. The function of the above process is expressed as: <h2 style=";text-align:left;direction:ltr">vec<h2 style=";text-align:left;direction:ltr"> 概念表征 <h2 style=";text-align:left;direction:ltr"> (a1,a2,…,a)<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr"> ] vec 年代表征 =[b1,b2,…,b n ] Among them, vec 概念表征 The embedding vector representation corresponding to the description text of the concept entity, vec 年代表征 Indicates the embedded vector representation of the description text of the year; q represents vec 概念表征 With vec 年代表征 The relationship strength between , n represents the dimension of the embedding vector model; vec 概念表征 ·vec 年代表征 Represents vec 概念表征 With vec 年代表征 The inner product operation, ||vec 概念表征 ||·||vec 年代表征 || represents vec 概念表征 With vec 年代表征 The Euclidean norm of .

6. The method for dating cave statues based on knowledge graph according to claim 5 is characterized in that: The establishment of multidimensional relationships in the concept map mentioned in S2.2 specifically refers to the establishment of multidimensional relationships between history, space, time, people and folk culture, among which: History: Combine time, events, and dynasty changes to reveal the evolution of statue styles; Space: Associate different regions and geographical relationships to analyze the distribution characteristics of cave statues; Time: Through the timeline positioning, the correlation between the image style and time is deduced; Characters: Combine statues, historical figures and their story backgrounds to enrich the content of the concept map; Folk culture: Integrate folk cultural characteristics and explain the interactive relationship between cave statues and folk culture.

7. The method for dating cave statues based on knowledge graph according to claim 6 is characterized in that: The S3 specifically includes the following contents: S3.

1. Dynamic relationship construction: integrating events, geographical space, and cultural customs to form a map of events and dynamically simulate the formation and dissemination process of cave statues; S3.

2. Comprehensive factor analysis: Integrate multi-dimensional information on characters, stories, space, and time to form a period model under a complex cultural background, and conduct in-depth exploration and analysis of regional characteristics, artistic style, folk culture, and historical value.

8. The method for dating cave statues based on knowledge graph according to claim 7 is characterized in that: The S4 specifically includes the following contents: S4.

1. Graph integration: Integrate the real physical world graph, concept graph and event graph to establish a multi-level association model and screen out key generation features; S4.2, Automated dating: Input the images of cave statues into the constructed multi-level correlation model to extract physical features; Map features to the knowledge graph to generate periodization results; the output periodization results include time, region, style features and explainability analysis; S4.

3. Use the knowledge graph to drive the dating process: Through the precise matching of statue images with nodes in the knowledge graph, multi-dimensional reasoning is performed in combination with time, space, and cultural characteristics to provide logical support for dating and cultural background explanation; S4.

4. Cross-dimensional correlation: Establish cross-dimensional correlation between physical world data and conceptual and rational layers. Focusing on the phenomenon of cultural integration, combine the artistic features and aesthetic styles of Buddhist statues from different dynasties to explore the deep correlation between time and space.

9. A knowledge graph-based cave statue dating analysis system for executing any of the methods described in claims 1-8.

10. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the instruction, program, code set or instruction set is loaded and executed by the processor to implement the cave statue dating analysis method based on knowledge graph as described in any one of claims 1-7.

Images