Sentiment analysis belongs to a classification technique with strong practical value, which can identify the hidden imagery in the text. In this paper, we study the sentiment tendency of words in Chinese language and literature texts, design a fuzzy ontology-based method for calculating textual sentiment tendency, build a textual sentiment tendency analysis model (SDNN) based on deep neural network and sentiment attention mechanism, and use the model in Chinese language and literature texts for empirical analysis. At 202 iterations, the Loss value of the SDNN model reaches the lowest value of 0.4626, and the corresponding accuracy, recall, and F1 values are 90.44%, 79.68%, and 84.72%, respectively, and tend to be stable. It indicates that the model has better classification performance and can be used in the task of analyzing the sentiment of Chinese language and literature texts. In addition, this paper explores the rhyme structure of character emotion vocabulary in Chinese language literary discourse, and takes Liu Zongyuan’s related works as an example for empirical analysis of the model. The results show that Liu Zongyuan’s emotion during his ten years in Yongzhou is generally in low, and the proportion of his negative emotion has been above 75%, which is reflected in the text vocabulary emotion is more sad than happy. This paper is of some significance for the study of the emotional tendency of vocabulary in Chinese language literary texts.

Sentiment analysis belongs to a classification technique with strong practical value, which can identify the hidden imagery in the text. In this paper, we study the sentiment tendency of words in Chinese language and literature texts, design a fuzzy ontology-based method for calculating textual sentiment tendency, build a textual sentiment tendency analysis model (SDNN) based on deep neural network and sentiment attention mechanism, and use the model in Chinese language and literature texts for empirical analysis. At 202 iterations, the Loss value of the SDNN model reaches the lowest value of 0.4626, and the corresponding accuracy, recall, and F1 values are 90.44%, 79.68%, and 84.72%, respectively, and tend to be stable. It indicates that the model has better classification performance and can be used in the task of analyzing the sentiment of Chinese language and literature texts. In addition, this paper explores the rhyme structure of character emotion vocabulary in Chinese language literary discourse, and takes Liu Zongyuan’s related works as an example for empirical analysis of the model. The results show that Liu Zongyuan’s emotion during his ten years in Yongzhou is generally in low, and the proportion of his negative emotion has been above 75%, which is reflected in the text vocabulary emotion is more sad than happy. This paper is of some significance for the study of the emotional tendency of vocabulary in Chinese language literary texts.

Sentiment analysis belongs to a classification technique with strong practical value, which can identify the hidden imagery in the text. In this paper, we study the sentiment tendency of words in Chinese language and literature texts, design a fuzzy ontology-based method for calculating textual sentiment tendency, build a textual sentiment tendency analysis model (SDNN) based on deep neural network and sentiment attention mechanism, and use the model in Chinese language and literature texts for empirical analysis. At 202 iterations, the Loss value of the SDNN model reaches the lowest value of 0.4626, and the corresponding accuracy, recall, and F1 values are 90.44%, 79.68%, and 84.72%, respectively, and tend to be stable. It indicates that the model has better classification performance and can be used in the task of analyzing the sentiment of Chinese language and literature texts. In addition, this paper explores the rhyme structure of character emotion vocabulary in Chinese language literary discourse, and takes Liu Zongyuan’s related works as an example for empirical analysis of the model. The results show that Liu Zongyuan’s emotion during his ten years in Yongzhou is generally in low, and the proportion of his negative emotion has been above 75%, which is reflected in the text vocabulary emotion is more sad than happy. This paper is of some significance for the study of the emotional tendency of vocabulary in Chinese language literary texts.

Traditional mechanical manufacturing experimental teaching is limited to one teacher demonstrating operations to several students at the same time, which is difficult to take into account and evaluate the differences in knowledge mastery of different students. In order to improve the above teaching defects, firstly, the teaching evaluation of students’ experimental level is carried out based on their experimental operation behaviors through K-means clustering. On this basis, a deep learning-based knowledge tracking SAFFKT model is designed to empower and update students’ knowledge status. A personalized teaching recommendation method for virtual simulation is proposed based on students’ knowledge state, and the hidden semantic matrix decomposition recommendation algorithm for teaching recommendation is improved and implemented. The AUC and ACC of SAFFKT model are significantly higher than that of the comparison model (p<0.01), and it is robust. The F1 value of the recommended experiments was 0.775, indicating a better recommendation effect. The teaching evaluation model achieves accurate classification of students' experimental behavior and yields different learning characteristics of three types of students. Therefore, the innovative work of virtual simulation teaching strategy in this paper is of practical significance.

In order to solve the enterprise data asset pricing problem in the digital economy environment, this paper utilizes machine learning algorithms such as multiple regression model, BP neural network, and random forest regression, respectively, to price enterprise data assets. Subsequently, the data obtained from each model is fused using the integrated Stacking algorithm to construct an enterprise data asset pricing model with integrated machine learning algorithms. Predictive estimation of the pricing of enterprise data assets is carried out after a detailed justification of the parameter selection of the model. The results show that data capacity, size, quality and freshness are the main influences on data asset pricing. The results of the parameter investigation show that the overall performance of the model is best when the number of node features is 7, at which time the explanatory degree and goodness of fit of the model are 94.33% and 97.27%, respectively. The accuracy, precision, recall and F1 value of the Stacking-based fusion model for enterprise data asset pricing prediction model increased by about 10% compared to the other three models, respectively, to achieve accurate pricing of enterprise data assets.

Teachers’ information literacy is related to the quality and efficiency of education and teaching in higher vocational colleges and universities. In this paper, a dynamic planning-based scheduling method is constructed to improve teachers’ time allocation efficiency and information literacy. First of all, according to the factors and constraints involved in the scheduling problem to determine the goal of solving the scheduling problem, mathematical model, and then the constraints involved in the scheduling of classes, converted into a dynamic planning of the mutually independent and related stages, with 1, 0 indicates whether to meet the constraints. By solving each stage and analyzing the solution of each stage, the optimal value function is summarized, and ACAA is used to traverse all the optimal solutions for each set of constraints. Examples are selected for scheduling test to verify the effectiveness of the algorithm, and the teacher information literacy assessment scale is designed. Applying the class scheduling algorithm to a higher vocational college, the mean value of the overall information literacy scores of the surveyed teachers is 0.15 points higher than the standard reference value, and the effectiveness of the class scheduling algorithm in this paper is verified. Practical experience (58.27%), teaching philosophy (50.19%), and subject requirements (33.36%) are the top three factors affecting teachers’ information literacy.

In the context of building an international consumption center city, it is of great significance to further study the competitiveness of the fashion industry and effectively grasp the direction and focus of the development of the fashion industry in order to promote the construction of an international consumption center city. The study adopts the entropy weight-TOPSIS method to measure the competitiveness of Tianjin’s fashion industry from 2020 to 2023, and compares it with typical provinces in order to have a comprehensive understanding of its fashion industry competitiveness level. Then, the spatial structure characteristics of the distribution of fashion industry facilities in Tianjin were further explored through the kernel density analysis method and the radius of gyration analysis method. Finally, Ripley’s K function is used to calculate the level of agglomeration and the range of the most significant agglomeration scale of each type of fashion industry, which summarizes the distribution characteristics of strategic fashion industries at the overall level. Horizontally, the competitiveness level of Tianjin’s fashion industry shows an upward trend from 2020 to 2023, and vertically, the competitiveness level of Tianjin’s fashion industry is ranked in the middle range of the country, with a certain gap between it and the strong provinces such as Jiangsu, Shandong and Guangdong. The most significant agglomeration scale of the new generation electronic information technology industry is 22,000 meters at maximum, and its DiffK value also reaches 13,317.938.

Piano timbre recognition and intelligent synthesis are of great significance in realizing the intelligent teaching of piano timbre. This paper takes the piano timbre teaching based on artificial intelligence interaction as the research object, constructs the timbre expression spectrum based on harmonic structure through the exploration of timbre synthesis, timbre features and other related theories, proposes the timbre feature extraction method based on the time-frequency cepstrum domain of the piano music signal, and then constructs the piano timbre recognition and intelligent synthesis system, realizes the simulation of the piano music, and then provides an intelligent interactive tool for the piano timbre teaching. The method is used to construct a piano tone recognition and intelligent synthesis system. When using the method in this paper, the amplitude of the piano tends to be stable when the frequency is 1600Hz~2400Hz, and there is no noise interference, and when the frequency is 2500Hz and 2800Hz, the amplitude is the lowest, and the recognition performance of the piano timbre is better. Meanwhile, the correct rate of timbre recognition of this method reaches 87.83%, which is better than 58.54% of the comparison method. In addition, the musical tone signals simulated by the method in this paper are very close to the theoretical values of each note of the real piano instrument captured, with an accuracy rate of up to 99%, which proves the accuracy of the simulated piano sounding. And the method can effectively promote the combination of artificial intelligence technology and piano teaching concept, the confidence level of quantitative regression analysis is high, and the evaluation results of teaching quality are good, which provides a reliable theoretical and practical basis for realizing the high-quality teaching of piano timbre.

The value assessment of ancient literary texts and the mining of linguistic features are indispensable parts of academic research and ancient cultural inheritance. This paper uses the multiple regression model as a quantitative analysis tool for value assessment to evaluate the value of ancient literary texts. At the same time, for the linguistic features of ancient literary texts, we put forward the quantitative descriptive definitions of words, phrases, sentences and other multi-layer and multi-latitude, and establish the corresponding calculation formulas. After the assessment of the value of ancient literary texts, it can be learned that, except for the artistic law and the breadth of dissemination, the ancient literary texts are positively correlated with other influencing factors such as the writing method and the rhythm and rhyme, and the gap between the predicted value of the value assessment and the real value is small, with an error of 40% or less in 90% of the cases. In the mining analysis of linguistic features using The Peony Pavilion and The West Wing as research objects, the average word length of the former is slightly higher than that of the latter, while the difference in the distribution of long and short sentences of the latter is relatively large. Meanwhile, the average dependency distance of The Peony Pavilion is 2.42, which is higher than that of The Story of the Western Wing by 0.1, making syntactic analysis more difficult.

Focusing on the learning behavior patterns of students with network behavior, this study mainly adopts sequence cluster analysis and lag sequence analysis to convert learning behaviors into sequences, and constructs a learning behavior pattern recognition model based on network behavior sequences. Aiming at different types of classroom learning behaviors in civic education under the network behavior sequence, a targeted teaching intervention mechanism is designed to help students convert their learning behavior patterns and thus improve their learning effects. In this paper, the online behaviors are clustered into four categories of “integrated, autonomous, compliant, and deviant” according to six level 1 codes, and the correlation coefficients of the online behaviors in the four learning categories range from 0.8539 to 0.9944, which is a very strong correlation. Finally, a survey of the results of the intervention in the classroom of Civic Education found that 75.22% of the students believed that the intervention had improved the learning effect of Civic Education. 67.7% and 77.54% of the students believed that the intervention had improved the enthusiasm and motivation of Civic Education learning. 79.04% of the students were willing to continue to learn independently according to the learning behavior pattern after the intervention.