Search

Reinforcement learning algorithms show significant variations in performance across different environments. Optimization for reinforcement learning thus becomes the major research task since the instability and unpredictability of the reinforcement learning algorithms have consistently hindered their generalization capabilities. In this study, we address this issue by optimizing the algorithm itself rather than environment-specific optimizations. We start by tackling the uncertainty caused by the mutual influence of original action interferences, aiming to enhance the overall performance. The Phasic Parallel-Network Policy (PPP), which is a deep reinforcement learning framework. It diverges from the traditional policy actor-critic method by grouping the action space based on action correlations. The PPP incorporates parallel network structures and combines network optimization strategies. With the assistance of the value network, the training process is divided into different specific stages, namely the Extra-group Policy Phase and the Inter-group Optimization Phase. PPP breaks through the traditional unit learning structure. The experimental results indicate that it not only optimizes training effectiveness but also reduces training steps, enhances sample efficiency, and significantly improves stability and generalization. [ABSTRACT FROM AUTHOR]

Deploying virtual machines poses a significant challenge for cloud data centers, requiring careful consideration of various objectives such as minimizing energy consumption, resource wastage, ensuring load balancing, and meeting service level agreements. While researchers have explored multi-objective methods to tackle virtual machine placement, evaluating potential solutions remains complex in such scenarios. In this paper, we introduce two novel multi-objective algorithms tailored to address this challenge. The VMPMFuzzyORL method employs reinforcement learning for virtual machine placement, with candidate solutions assessed using a fuzzy system. While practical, incorporating fuzzy systems introduces notable runtime overhead. To mitigate this, we propose MRRL, an alternative approach involving initial virtual machine clustering using the k-means algorithm, followed by optimized placement utilizing a customized reinforcement learning strategy with multiple reward signals. Extensive simulations highlight the significant advantages of these approaches over existing techniques, particularly energy efficiency, resource utilization, load balancing, and overall execution time. [ABSTRACT FROM AUTHOR]

The ecological psychologist James J. Gibson defined the notion of affordances to refer to what action possibilities environments offer to animals. In this paper, we show how (artificial) agents can discover and exploit affordances in a Multi-Agent System (MAS) environment to achieve their goals. To indicate to agents what affordances are present in their environment and whether it is likely that these may help the agents to achieve their objectives, the environment may expose signifiers while taking into account the current situation of the environment and of the agent. On this basis, we define a Signifier Exposure Mechanism that is used by the environment to compute which signifiers should be exposed to agents in order to permit agents to only perceive information about affordances that are likely to be relevant to them, and thereby increase their interaction efficiency. If this is successful, agents can interact with partially observable environments more efficiently because the signifiers indicate the affordances they can exploit towards given purposes. Signifiers thereby facilitate the exploration and the exploitation of MAS environments. Implementations of signifiers and of the Signifier Exposure Mechanism are presented within the context of a Hypermedia Multi-Agent System, and the utility of this approach is presented through the development of a scenario. [ABSTRACT FROM AUTHOR]

The Bdi model of rational agency has been studied for over three decades. Many robust multiagent systems have been developed, and a number of Bdi logics have been studied. Following this intensive development phase, the importance of integrating Bdi models with inconsistency handling and revision theory have been emphasized. There is also a demand for a tighter connection between Bdi-based implementations and Bdi logics. In this paper, we address these postulates by introducing a novel, paraconsistent logical Bdi model close to implementation, with building blocks that can be represented as Sql/rule-based databases. Importantly, tractability is achieved by reasoning as querying. This stands in a sharp contrast to the high complexity of known Bdi logics. We also extend belief shadowing, a shallow and lightweight alternative to deep and computationally demanding belief revision, to encompass agents' motivational attitudes. [ABSTRACT FROM AUTHOR]

Digital Twins (DTs) are emerging as a fundamental brick of engineering Cyber-Physical Systems (CPSs), but their notion is still mostly bound to specific business domains (e.g. manufacturing), goals (e.g. product design), or applications (e.g. the Internet of Things). As such, their value as general purpose engineering abstractions is yet to be fully revealed. In this paper, we relate DTs with agents and multiagent systems, as the latter are arguably the most rich abstractions available for the engineering of complex socio-technical and CPSs, and the former could both fill in some gaps in agent-oriented software engineering and benefit from an agent-oriented interpretation—in a cross-fertilisation journey. [ABSTRACT FROM AUTHOR]

Negotiation is a key form of interaction in multi-agent systems. Negotiation enables agents to come to a mutual agreement about a goal or plan of action. Current negotiation approaches use traditional interaction protocols which do not capture the normative meaning of interactions and often restrict agent autonomy. These traditional negotiation approaches also have difficulty specifying accountability. This paper seeks to address this gap in maintaining autonomy and capturing accountability during negotiation through the use of normative commitments. We propose Nala, a commitment-based negotiation semantics. Nala uses commitments to provide normative meaning to agent interactions. The nature of commitments support in capturing accountability through the violation of created commitments. We illustrate Nala's usage via a case study using a game scenario where agents participate in negotiation to bring about their goals in a research constrained environment. [ABSTRACT FROM AUTHOR]

Algorithmic recommendation is an important means of using multimedia technology for cultural dissemination and helping rural cultural revitalization. In order to realize the effective dissemination of Luo Yue rice culture, this paper constructs a hybrid recommendation algorithm that recommends videos about Luo Yue rice culture according to user preferences, which uses the cultural content relationship derived from LDA to replace the recommendation list generated by collaborative filtering and crosses it with the historical rating recommendation list to solve the problems of recommendation accuracy and cold start. The analysis of the dissemination effect shows that the average MAE value of this paper’s algorithm is 0.671, the proportion of users who watch relevant videos “occasionally” and more frequently has increased by 5.84%, and users recognize the recommendation method. The algorithm proposed in this paper has a good recommendation performance and helps to disseminate rural culture, as demonstrated by the above results.

This paper mainly discusses the combination of 3D modeling technology and traditional characteristics of ink animation to promote the innovation and development of Chinese traditional animation. Aiming at the characteristics of ink painting, the 3D model construction algorithm of ink animation is proposed, which realizes the establishment of algorithms such as transforming the 2D model into a 3D model, constructing the 3D animation scene, imitating the virtual reality of ink painting and rendering the virtual reality of ink animation. The basic model of the particle system is proposed, including modules for particle generation, particle properties, particle motion change, particle extinction, and particle drawing. Combined with the basic principles based on the particle system model, randomly distributed particles break the regular rectangular shape in the overall distribution. More realistically simulates a variety of different animation forms for ink animation. The simulation speed and convergence time of this paper’s model are better than the other two models. The simulation results fulfill the demands of real-time ink animation performance. The FPS of the most complex animation image, the color ink staining of the cow, is 31FPS, which also meets the requirements of the animation image. It shows that the combination of 3D modeling technology, particle system-based dynamic simulation, and ink animation is perfect and has a broad application prospect.

The number of college graduates is increasing year by year, and college students’ career planning and employment have always been a high concern of the community. The article describes the specific role of artificial intelligence technology in assisting college students’ career planning and employment guidance and combines WEB technology to construct a college student’s career planning and employment guidance service platform. The information entropy is used to construct the feature importance set of student employment data, and the simple Bayesian algorithm is used to realize the construction of the classification model of student employment service data in combination with the feature importance. On the basis of considering the selection of students’ employment information, the time decay factor is introduced to improve the personalized recommendation algorithm, and the cosine similarity between students’ employment positions is calculated so as to realize the generation of students’ employment service information recommendation. The area of the classification ROC curve of the plain Bayesian model is more than 95%, the classification time for student employment service data is only 5.62ms, and the F1 value of the employment service information recommendation result is 6.03% higher than that of the traditional method. The college students’ career planning and employment guidance service platform was only satisfactory for 31.77% of the students, with the highest satisfaction score of 3.69 for getting information. Relying on artificial intelligence technology can help students make personalized recommendations for employment service information, which can improve the level of career planning and guidance for college students.

The inheritance and protection of the historical, cultural and folk art values of ancient villages in development are generally carried out in three aspects, namely, architectural elements, landscape elements and living elements. This paper applies a 3D animation scanner to collect the point cloud data of Guibei ancient villages and completes the pre-processing of the point cloud data by using the algorithm of SAC-IA coarse alignment + ICP fine alignment for the alignment. After that, the Poisson equation is optimized by using Gaussian filtering to realize the point cloud surface reconstruction of Guibei village buildings, and the cultural information texture extracted by the K-Means clustering algorithm is mapped to realize the digital reproduction of the cultural heritage of ancient villages in the Guibei. On this basis, the cultural brand image of Guibei ancient villages is constructed from multiple perspectives by collecting tourists’ cognitive data about them. The accuracy of the established 3D animation model is as low as 88.02% and as high as 97.21%, the total fluctuation of completeness ranges from 84.93% to 96.43%, and the time used for modeling basically stays below 2.9h. It has a high awareness of its natural landscape (26.44%), humanistic landscape (37.33%), folk culture (20.57%), service facilities (9.23%), characters (4.40%), and specialty food (2.03%). The study provides a theoretical and practical foundation for communicating images of ancient villages.

This paper gathers information on artificial intelligence, art communication, news communication, and other related topics, with the aim of comprehending the role of artificial intelligence in enhancing information and new media art communication. The paper selects information from the “Annual Development Report of China’s News Industry” and “Review of China’s Advertising Market” to analyze and summarize the current state of the gradual intelligentization of news. We examine the browsing data from new-era news media, including today’s headlines, to investigate the impact of intelligent news dissemination. Lastly, we scrutinize the spatial data from actual new media artworks to comprehend the role of intelligence as a catalyst for experiencing a sense of immersion. The results show that by June 2024, the number of Internet users will be as high as 725 million, accounting for 96.8%. The penetration rate of intelligent new media also rises gradually with changes in time. It reached 84.9% in 2023, and almost half of the users are undergoing intelligent migration. And Tencent News, as a new-age news media, tops the list with a monthly activity of 27,042.17. The interactive nature of artificial intelligence technology promotes artistic communication and broadens the scope of communication.

With the rapid development of artificial intelligence technology, colleges and universities pay more and more attention to innovation and entrepreneurship education. The path of artificial intelligence technology is proposed to empower the high-quality development of innovation and entrepreneurship education in colleges and universities from multiple perspectives. An optimization strategy is proposed for an innovation and entrepreneurship education resource-sharing platform based on a collaborative filtering algorithm. The proposal is for a collaborative filtering recommendation algorithm (CFUR) that is based on user feature weight confidence. Finally, experiments have shown that the EP value of the improved algorithm in this paper is significantly higher than that of other recommendation algorithms. Through teaching practice, students’ satisfaction with the educational resource-sharing platform based on the improved algorithm outperforms that of the traditional teaching method, and the mean value of test scores for innovation and entrepreneurship ability surpasses that of the traditional teaching class by 17.07. It shows that the optimized resource-sharing platform has a significant impact on students’ innovation and entrepreneurship abilities, which helps to cultivate innovative and entrepreneurial talents.

Based on the theoretical knowledge of woodworking skills and digital restoration, this paper designs a questionnaire on traditional woodworking skills and digital restoration technology, conducts a reliability test on the questionnaire, adopts online and offline methods to distribute the questionnaire and obtains research data. Pearson’s coefficient and multiple linear regression model were used to explore the relationship between traditional woodworking skills and digital restoration technology from the perspective of non-heritage. The results show that the regression equation of the historical value of woodworking skills = 0.428+0.176*Efficiency Improvement+0.208*Quality Optimization+0.235*Talent Cultivation, and the explanatory ability of Efficiency Improvement, Quality Optimization (X2), and Talent Cultivation (X3) for the historical value of the skills is 76.6%, and there is no problem of multicollinearity. The mechanism of interaction between traditional woodworking techniques and digital restoration technology is analyzed from the perspective of non-heritage, which is important for the preservation and dissemination of non-heritage.

This paper applies artificial intelligence technology to libraries to promote service innovation in information resource construction, intelligent push and information analysis. Starting from college libraries, we collect relevant news reports and activity message texts from their libraries as textual materials for analysis and utilize the LDA subject model to reveal the current development status of the application of cultural inheritance in smart college libraries. A questionnaire was used to explore the tendency of teachers and students to promote cultural resources through RSS in smart college libraries. The results indicate that in smart libraries’ cultural heritage work, online services and knowledge resource services are the primary areas of focus, with their thematic strengths being 0.13874 and 0.12912, respectively. As time goes by, the intensity of the cultural theme of red gradually increases. The trend of RSS pushing resources in red culture, morality, philosophy, art, and religion is more than 60%.

Multi-dimensional information fusion technology helps to realize the synergy and complementarity between the information, which provides a guarantee for the settlement monitoring and early warning of electric power pole towers. The settlement monitoring algorithm for electric power pole towers is improved by this paper based on this technology and a settlement warning model is proposed. The traditional PDA algorithm can be enhanced by enhancing the pole tower image acquisition capacity with the wavelet transform algorithm and achieving the fusion and evaluation of multidimensional information through an isolated forest algorithm. The relationship functions of tilt angle, settlement wind speed size and other indicators are established, and the system of super-definite equations is used to solve the correlation coefficients, and the early warning system of pole tower settlement is constructed. The analysis results show that the monitoring results of the monitoring algorithm are fluctuating and stable; the absolute value of the error in the horizontal direction is not more than 9 mm, and the absolute value of the error in the vertical direction is not more than 14 mm. The predicted values of the maximum displacement and stress R2 are close to 1, and the MAPE is 0.436% and 1.123%, respectively. It indicates that the improved power pole tower settlement monitoring algorithm and early warning system in this paper have satisfactory performance.

In the field of apparel design, the traditional manual work method can no longer meet the requirements of production and market, and the application of information technology in textile and apparel design has become increasingly important. The article outlines the significance and value of integrating cultural elements in textile apparel design and evaluates the significance and applications of information technology in textile apparel design. From the perspective of cultural genes, the article explores the cultural design factors in textile apparel design and establishes a cultural gene pool for textile apparel design. Based on the DCGAN model, the DCGAN model is improved by optimizing the loss function and residual module to construct the textile clothing style migration model. In order to verify the effectiveness of the above model in the sustainable design of textile apparel, the style migration of textile apparel and visual similarity were verified, and the attributes of sustainable design needs, importance, and satisfaction of textile apparel were evaluated. The IS values of the improved DCGAN model for textile apparel design were 0.428 and 0.439, and the visual similarity scores of the sample and migrated apparel ranged from 6.21 to 7.03. The coefficient of incorporating cultural themes reached a maximum of 75.19%, and the subjects’ satisfaction score with the culture of the textile dresses was the highest at 6.23 points. Relying on information technology can explore the connotations of cultural elements and provide a diversity of textile apparel design development.

We determine the content of natural scenery material in landscape painting based on the artistic characteristics and techniques of traditional Chinese landscape paintings. The 3Dmax software in 3D modeling technology is selected as the main modeling tool, and the construction of the 3D scene of landscape painting is completed based on the IFS theory (Iterative Function System Theory), and then the scene is rendered and texture synthesized, and the virtual landscape mood space is displayed, and the fusion of natural scenery and 3D modeling technology is finally realized. The analysis of the research data reveals the integration of natural scenery and 3D modeling technology. The results show that the iterative function system quickly generates the natural scenery scene in the landscape painting by continuously regulating, changing, and controlling the parameters in the affine transformation (the influence factor of each branch is 0.9, 0.7, 0.4, and 0.15). While the static rendering speed of 47 is slower than the dynamic rendering speed of 34, the overall rendering speed still meets the real-time detection criteria, confirming that 3Dmax is capable of achieving dynamic real-time rendering of 3D mountain models. The regression coefficient for high efficiency is 0.208, with a significance level of 0.005, indicating a significant positive impact on user experience. The higher the efficiency, the better the user experience, revealing the mechanism behind the interactive 3D scene. This study promotes the integration of natural scenery and 3D modeling technology to further develop the digital research of Chinese landscape painting.

The current inheritance of traditional handicrafts has ushered in unprecedented historical opportunities and challenges. This paper explains the process of designing handicraft products that are not heritage-based by describing traditional handicraft products. Using image processing technology, the collected images are pre-processed, and after extracting the image features based on a clustering algorithm, the features are reorganized, deduced, and fused using shape grammar to achieve the purpose of NRH’s innovative design. The Kano model is used to analyze the demand for non-heritage innovation design based on shape grammar. The top 5 user demand items of importance for NRH product innovation design are durability (0.7982) > utility (0.7906) > ethnic difference (0.7887) > regional culture (0.7223) > personalized design (0.7507). The combination of the Kano model and information technology enables precise positioning of the entire process of NRH innovation design.

Clinical psychological counseling suffers from problems such as a shortage of medical personnel and uneven quality, and artificial intelligence technology provides a feasible way to solve these problems. In this paper, a chatbot model for psychological counseling is designed using Seq2seq, and Encoder-Decoder and Attention mechanisms are introduced to improve decoding accuracy. LSTM is used as the basic unit, and the beam search algorithm is added to improve the diversity of replies. The experimental results show that adding LSTM and Beam Search can generate higher-quality and more natural psychological counseling responses, and the loss value of this paper’s model decreases to 1.33 after 10 rounds of training. The total score of the OQ-45.2 questionnaire of the experimental group’s post-test decreased by 13.8 points, and the mean value of symptom distress decreased by 8.26, performing significantly better than that of the control group. The chatbot design in this paper is reasonable and aids in improving the quality of clinical psychological counseling services.

With the development of artificial intelligence, the use of generative adversarial networks in deep learning can generate good visual art based on content, and at the same time, it can reduce the problems of noisy texture, transition migration, and image distortion that exist in content-based image generation. The article proposes a multi-view image generation architecture (DrawGAN) based on generative adversarial networks and builds an AI design tool on this basis to explore the effect and analysis of the design tool for content generation based on generative adversarial networks on visual art. Through experimental testing, the model proposed in this paper, after the constraints of the relevant loss function, the generated image content is compatible with the input image, the style is consistent with the style of the target image, and the image generated by the base DrawGAN has a high degree of reproducibility. In addition, the average score of visual art images generated using the model proposed in this paper is 8.25, which is 2.26 points higher than that of images generated by traditional methods. To sum up, the DrawGAN model proposed in this paper is a valuable tool in the generation of visual art images.

The low-carbon upgrading of transmission grids can be effectively realized by trend optimization techniques. This paper improves the traditional trend calculation based on the distributed power supply controlled by droops. Each node’s power balance equations are outlined based on the established distributed power supply with droop control and other models. The improved Niu La method is used to establish the trend calculation model, and the active sag coefficient and reactive sag coefficient in the droop control are normalized. Virtual impedance is utilized to regulate the reactive power classification between distributed power sources in the transmission network to minimize transmission network losses. The carbon and economic benefits of this paper’s method are analyzed using the MATLAB platform, and the carbon emissions of line and node active losses are reduced by 72.30% and 50.67%, respectively, before and after the optimal flow control. The economic benefits of saving line losses after performing optimization and control are 4.8739 × 105 yuan, 7.8593 × 105 yuan and 1.1283 × 106 yuan, respectively.

With the deepening of economic globalization, countries around the world have put forward new requirements for talent. College students, being a unique segment of society, must naturally adapt to these significant changes and undergo new enhancements. This paper utilizes undergraduates from a university as the research subject and develops a leadership model for college students by designing specific research variables. We use the questionnaire survey data to empirically investigate the relationship between the antecedent variables of college students’ leadership and their influence. To compare the means of the four antecedent variables, descriptive statistics were used. The levels of social participation were better than medium or above (p > 3.5), while the levels of organizational factors were lower and needed to be improved. Correlation and regression analyses verified the relationship between each dimension of the four antecedent variables and college student leadership. Some aspects of college students’ leadership are associated with the dimensions of social participation, family function, personal endowment, and organizational factors. Among them, the degree of social participation in the dimension of “participation in public affairs” will affect the development of college students’ leadership, and the correlation coefficient of the two is 0.663, which is positively correlated at the 0.01 confidence level. The mediation effect test concludes that social value mediates the influence of college students’ social participation in leadership. Finally, the research results propose a strategy for volunteering that uses artificial intelligence to help college students develop their leadership skills.

Blockchain, as a secure distributed ledger technology, has received extensive attention from academia and industry for its decentralization, comparability and traceability. This paper proposes a method for verifying data security on blockchain, and a cryptographic uploading and retrieval scheme is combined with IOTA blockchain technology. In this way, an IoT data storage and sharing model is designed to combine blockchain with IPFS. In this paper, it is found that there is a certain delay in operating on Fabric and IOTA compared to the scenarios that do not use blockchain. But Fabric has better performance in upload and download performance than IOTA. In the proposed scheme in this paper, the uploading time is stabilized between 0.8 and 1.2, and the retrieval time is stabilized between 0.43s and 0.76s, which has superior data uploading and retrieval performance than Scheme 1. Furthermore, the time to generate a key for a 50MB file is only around 1800ms, and for a 50MB file split into 6KB blocks, the time to generate a secret key is only around 36s. In this paper, when the file size reaches 5MB, the computational overhead of encryption and decryption is about 1315ms and 1283ms, respectively, and the time overhead of data consistency validation is only about 1095ms when the data block is cut into 6KB for a file of 5MB, indicating that this paper’s scheme has a lower time overhead in the generation of the secret key, encryption and decryption, and the data consistency validation. In summary, the above experiments verify the security and applicability of this paper’s scheme in data storage and verification.

Language and society are a pair of covariance; language reflects the changes in society, and the changes in society have an important impact on the language. The urbanization process in the Subei area is reflected in the current Mongolian language use situation. This paper explores the intrinsic connection between language and the Chinese national community, in which the emotional identity of Mongolian script is one of the factors affecting the development of Mongolian language in Subei. Mongolian emotions are represented semantically using the SOFT self-attention mechanism. The computation of emotion word weights and distributions is carried out sequentially to enhance the emotion word tokens that semantic rules represent. Through the mining of emotion prior knowledge, new emotion words are found from the Mongolian emotion corpus with the help of emotion seed words, and a set of candidate emotion words is constructed. The emotion words that were mined in Mongolia are merged into emotion sentences, and then they are emotionally masked. An empirical survey is designed to visit and examine the local Mongolian residents in Subei using random sampling. The psychological distance of Mongolians towards intermarriage is the lowest among the subject ethnic groups, with a mean value of 1.7585. The development of the Mongolian language in Subei is a map of the progress of the Chinese nation’s consciousness and contributes to the formation of a sense of community among its people.

The development and application of machine translation cannot be separated from the evaluation of its quality. Correct evaluation is a correct lead to its development direction, and the translation direction respected by mainstream evaluation can be said to be the development direction of machine translation in research. In this paper, the Transformer model of the multi-head attention mechanism is constructed, and the optimization and improvement schemes in both model architecture and training data are proposed for the mBERT and XLM models to form the XLM-R model. Based on the predictor-evaluator QE model and the pre-trained model-evaluator QE two models, the XLM-R translation quality assessment baseline is proposed. A CLS pooling method is proposed for the baseline system, which ensures similarity at a finer-grained subword level. Meanwhile, the subword similarity scoring index is embodied using a simple splicing method, taking into account the global semantic quality. Simulation experiments are set up to evaluate the quality of translation tasks at the sentence and system levels, respectively. There was a 7.92% improvement in the Pearson correlation coefficient of the model with similarity characterization when compared to the one without the model in the sentence-level QE task development set, which was 0.0379 higher. The model’s translation fluency also showed an increasing trend in the interval of 0.15 to 0.5 with the growth of the manual score, indicating that the model proposed in this paper is effective in baseline system evaluation.

With the development of science and technology, the ocean is providing a strong material and economic guarantee for the development of human society. The issue of assessing the health of the marine ecosystem and providing data support for the healthy development of the marine system has become a hot topic for many scholars. This study mainly establishes a marine ecosystem health assessment system, analyzes the health degree of marine ecosystems using information entropy, and then adopts the gray correlation method and the DPSIR model to predict the health level of marine ecosystems, discusses the sustainability of the marine generating system, and provides a development direction for the healthy development of marine ecology. The entropy of the “supportive input” criterion layer is the highest, accounting for 42.25%. From the analysis of the comprehensive health index from 2020 to 2023, the IHI ranges from 0.6 to 1.5, and the state of the ocean can reach a healthy level in the future period, which shows that the development of the economy and science and technology influences the healthy development of marine ecology.

Starting from the basic principles of talent cultivation mode, this paper proposes a reform strategy to optimize the talent cultivation mode of environmental design with the help of intelligent technology in order to improve the quality of talent cultivation and achieve the “first-class” level for the current talent cultivation mode of environmental design majors. The necessity of an intelligent database in environmental design teaching is analyzed, and the improved TF-IDF algorithm and automatic classification algorithm of resources are proposed to improve the search for teaching resources for environmental design majors. Then, based on the query, the teaching resource recommendation model (UBM) is proposed. An algorithmic comparison between the improved TF-IDF method and the UBM model is carried out to verify the effectiveness of the improved algorithmic model. Combining the traditional talent training model results with the intelligent technology-assisted talent training model results for independent samples of the T-value test. T-value test results show that the F-value is 0.317, the Sig-value is 0.782>0.05, and the significance corresponding to the ANOVA t-test value is 0.009

Hainan’s marine intangible cultural heritage, with its unique value, occupies an important position in global cultural diversity, but it is currently facing an endangered situation. This paper utilizes the Delphi method to select the initial indicators for the evaluation of the endangerment of Hainan’s marine intangible cultural heritage, then applies the hierarchical analysis method to determine the weights of the indicators at each level and evaluates the endangerment of Hainan’s marine intangible cultural heritage based on the fuzzy comprehensive evaluation method. Then, according to the current situation and evaluation results of Hainan marine intangible cultural heritage, the digital inheritance and protection path of Hainan marine intangible cultural heritage is designed by using modern technology as a means. Among the specific indicators of the endangerment evaluation index system of Hainan marine intangible cultural heritage, the economic benefit has the largest weight, which is as high as 0.104. It shows that the economic benefit of preserving and preserving Hainan marine intangible cultural heritage needs to be improved. In the endangerment evaluation of Hainan’s marine intangible cultural heritage, for example, the overall evaluation value is 3.525, and its endangerment status is between generally endangered and relatively endangered. Among them, the evaluation values of social and cultural factors are higher, favoring comparatively endangered, indicating that more consideration should be given to the inheritance and protection of A Hainan marine intangible cultural heritage from both social and cultural aspects.

We propose the AI painting model based on the generative adversarial network and analyze the color, pattern, and visual hierarchy of the AI painting’s visual expressiveness. Starting from the painting sketch, the corresponding generator and discriminator are integrated to construct the AI painting model based on a generative adversarial network, and the training strategy, optimization objective, and loss function calculation formula of the model are given. The visual analysis of the AI painting based on deep learning technology is combined with research analysis. In terms of simplifying sketches and generating pictures, the model presented in this paper performs more prominently than the control models (CNN and SVM), which is reflected in the loss value and F1 value, respectively. In addition, there are significant differences between AI paintings and traditional paintings in the use of color (P = 0.035, T = 1.478), pattern cultural allegory (P = 0.011, T = 0.347), and visual hierarchy (P = 0.008, T = 2.993), and AI paintings are more visually expressive than traditional paintings, and at the same time, the regression equation of the AI paintings’ visual expressiveness is 0.125+0.179*Color Use+0.188*Pattern Cultural Allegory+0.234 Visual Hierarchy. This study provides a clearer perception of the modern art creation trend and its visual expressiveness, which is of enormous significance in boosting the development of the art field.

Music resource-sharing platforms can increase students’ employment paths and enhance their employment ability. Load prediction and resource scheduling algorithms are designed in this paper to construct a music resource-sharing platform based on cloud computing technology. The mean predictor’s predictions are combined with the GRU-CNN model’s predictions, and the weighted average is calculated to enhance the accuracy and stability of the predictions. Then, the integrated migration scheduling method is designed to make the load of the cloud environment more balanced and the resource utilization more reasonable by the four processes of migration triggering, container selection, node selection, and container migration in a coarse-grained dynamic scheduling approach with nodes as objects. Following the completion of the construction, the algorithm that was designed is evaluated for load prediction and resource scheduling, and regression analysis is performed to examine the role of music resource-sharing platforms in promoting employment and entrepreneurship. It is found that after the resource scheduling by the algorithm in this paper, the node operation tension is greatly alleviated, and the CPU resource utilization and memory resource utilization of node Node1 are reduced from 80.96% and 62.69% to 58.03% and 43.36%, respectively. The resources of the low-load node Node4 are effectively utilized, the memory resource usage rate has increased from 6.86% to 49.52%, and the resource usage of each node in the cluster is more balanced and reasonable. At a statistical level of 1%, the impact of music resource-sharing platform development on employment quality is 0.194. In contrast, the industrial structure does not significantly contribute to the quality of employment, indicating that under the current economic situation, emerging economic forms such as music resource-sharing platforms are the main driving force to improve the quality of employment and the development of music resource sharing platforms generally improves the quality of labor force employment.

NPMI is used to describe the local co-occurrence properties of words, and the collected text data is constructed into a homogeneous graph composed of words. A word-granularity graph convolutional neural network (VGCN) is used as a graph encoder to achieve parallel training of graph convolutions. The output of the VGCN is used as a kind of embedding vector for BERT, and the VGCN is fused with other embedding using a transformer for feature fusion. The reprojection layer converts the model output to semantic representation vectors, and the post-projection layer converts the semantic representation vectors to vectors for loss computation. NT-Xent is chosen as the loss function of the model, while the real-time semantic understanding effect of the model is evaluated using the Spearman correlation coefficient in statistics. The CBOW model provided by Word2vec is learned from the training set and constantly updated for each parameter to calculate semantic similarity between words. The LDA model is used to extract topic keywords from each document and calculate the probability, which finally completes the dynamic assessment of the credibility of text content based on semantic understanding. The ROUGE-L index and BLEU-4 index of the VGCN-CL model in this paper are 66.47% and 63.86%, respectively, which are 23.20% higher than the baseline model ROUGE and 25.14% higher than BLEU-4. When the content assessment results are true, the accuracy and other metrics of this paper’s model are 90.36%, 89.87%, and 90.09%, respectively, which are the best performance among all models.

Image recognition technology is widely used in the field of criminal investigation or in the civil field, in which fingerprint image recognition technology, as a means of verification, has a very high legal authority and plays an irreplaceable role. In this paper, image recognition technology is specifically applied to the extraction of fingerprint features from crime scenes. The proposed fingerprint image curve changes are used in the image patch recognition algorithm. By analyzing the changes in the fingerprint curve profile, the evolution law of the fingerprint curve is calculated. Then, using the discrete iteration method, the missing fingerprint image is repaired for recognition. Experiments show that the method in this paper can effectively repair the missing fingerprint image part. Despite the low recognition rate for fingerprint images of 88.1%, other quality images are correctly recognized at more than 95% and have a relatively good recognition effect. Legal practitioners are most satisfied with the accuracy of feature extraction using this technique after using it.