With the rapid development of big data and AI technology, programming is in high demand and has become an essential skill for students. Meanwhile, researchers also focus on boosting the online judging system's guidance ability to reduce students' dropout rates. Previous studies mainly targeted at enhancing learner engagement on online platforms by providing personalized recommendations. However, two significant challenges still need to be addressed in programming: C1) how to recognize complex programming behaviors; C2) how to capture intrinsic learning patterns that align with the actual learning process. To fill these gaps, in this paper, we propose a novel model called Programming Exercise Recommender with Learning Style (PERS), which simulates learners' intricate programming behaviors. Specifically, since programming is an iterative and trial-and-error process, we first introduce a positional encoding and a differentiating module to capture the changes of consecutive code submissions (which addresses C1). To better profile programming behaviors, we extend the Felder-Silverman learning style model, a classical pedagogical theory, to perceive intrinsic programming patterns. Based on this, we align three latent vectors to record and update programming ability, processing style, and understanding style, respectively (which addresses C2). We perform extensive experiments on two real-world datasets to verify the rationality of modeling programming learning styles and the effectiveness of PERS for personalized programming guidance.
The rapid growth of Massive Open Online Courses (MOOCs) has revolutionized the education landscape by providing accessible and flexible learning opportunities. However, with the emergence of hybrid learning models, how to integrate MOOC into traditional classroom environments and effectively utilize MOOC video content in face-to-face teaching has become a current challenge. To solve this problem, we propose an instructive video locating system for MOOC hybrid teaching. The system analyzes teacher instructions and leverages technologies such as speech recognition, text segmentation, and natural language processing to easily locate relevant video clips in MOOC courses, enhancing the blended teaching experience. We conducted experiments using educational videos in MOOC online courses. The results show that the system can accurately locate video clips based on text queries. Compared with manual searches, the accuracy rate exceeds 85
This paper selected an international school in Brunei as a research case, and used the Learning Behaviors Scale (LBS) to analyze the learning behaviors of learners in Chinese learning. According to statistics gathered by questionnaires, this essay analyzed the results via SPSS 22.0, and reported that when learning Chinese as a second language, all the four dimensions of learning motivation, attitude, attention and strategy were positively correlated with the learners’ academic performance; differences in learning duration could lead to those in learning behaviors; differences also occurred in the learning behaviors between male and female students, i.e. boys performed weaker than girls in terms of the above four dimensions; there was a correlation between learners’ language learning behaviors and their family languages, i.e. different family languages had an impact on all the four dimensions of learners’ learning behaviors, among which difference between motivation and strategy was quite big; in addition, differences in religious beliefs did not affect learners’ second language learning behaviors. Based on these research results, the paper advocates that the idea of computer-based education can be introduced to help students establish reasonable Chinese learning behaviors. At the same time, this study can provide several useful tips on international Chinese education integrated with computer technology.
The packet transport network (PTN) is becoming increasingly popular in mobile communication as a technology to provide efficient transmission. With the dramatic surge in user numbers, the PTN must be able to carry more. However, the existing PTN has low resource utilization and poor network security, Therefore, optimizing every aspect of the current PTN is essential. For optimizing PTN network, the decision of both carrier users and service product suppliers should be considered. When satisfying the decision of the product supplier, it is necessary to consider a number of evaluation indicators in the PTN, and there may be some correlation between the indicators. Hence, This study introduces a multi-objective optimization approach inspired by the Gray Wolf Algorithm and centered around the optimization of a PTN network. Carrier user is taken as the upper decision maker, and the goal is to pay the supplier with the lowest possible cost. Suppliers play a lower role in decision-making, considering the PTN Optimal performance, mainly including two goals, the first goal is to have the highest evaluation score of the LSPOR, and troubleshoot the label switched path (LSP) anomaly on the network. As a secondary objective, By maximizing chain bandwidth utilization ratio (CBWUR), we can resolve the problem of high committed information rate (CIR) bandwidth usage. Results show that the model allows services to use shorter paths, which improves resource usage and reduces costs for upper-level decision-makers. At the same time, the model can also improve the evaluation scores of LSPOR and CBWUR indicators, which improves the security of the PTN, reduces the cost of additional network resources, and optimizes the performance of the PTN to the greatest extent. This paper primarily focuses on the following topics:
The online data science learning platform offers students an adaptable educational experience and a versatile learning and experimental environment. However, with the escalation of system concurrency, performance concerns including resource limitations, protracted online experiment response, and platform breakdowns, all of which can have a detrimental effect on the advancement of data science experimental learning. This paper proposes a performance analysis system for online data science learning platforms. It uses the Application Performance Index (Apdex) to evaluate user satisfaction of system performance, and promptly detects real-time performance anomalies and forecasts performance trends by LightGBM and LSTM algorithm. Meanwhile, the performance bottlenecks are also identified and correlated based on the function invocation chains with various time series performance metrics data. Therefore, negative user experience is effectively prevented. Furthermore, we conducted a series of experiments based on 40 days of operational data from a real online data science learning platform, the effectiveness and superiority of the proposed method in this system are clearly verified.
