Previous lectures targeted at overseas Chinese children have mostly been offline in the form of youth camps sponsored by the Chinese governments before the COVID-19 pandemic and are now usually administered online in a single-topic format transnationally post-pandemic. This form of “one-to-many” singular online lectures comes along with discontinuity, mass audiences, chaos, and being less tailored to the individual needs and context, failing to meet the evaluation metrics in different evaluation levels. This paper proposes a model of transnational online livestreaming serial socio-cultural lectures for overseas Chinese children, which puts forward the synergy of lecturing staff from higher educational institutions (HEIs) in China and students at Chinese language schools (CLSs) worldwide as participants. Lecturing staff from needs and context HEIs can guarantee sufficient keynote speakers for “one-on-one” mode and cope with the above-mentioned problems. After the implementation of five sessions of these serial lectures, evaluation of this model shows that the pre-lecture, during-lecture, and post-lecture stages have are applicable. Furthermore, for sociocultural lectures, the audience may be more interested in literature classics than folk arts from a specific region. Correlation analysis reveals that younger children have a better evaluation of this model and an improved inclination to attend lectures, which sheds light on the improvement of sociocultural lectures as online lectures are preferred among overseas Chinese children.

While college English teaching is steadily changing from static knowledge transfer to dynamic language ability development, classroom activities centered on language application are becoming more and more important in cultivating students’ language application ability. In recent years, education has been paid more and more attention, the scale of university education has gradually expanded, the professional categories have become more and more complete, the curriculum has become larger and larger, and the number of students has grown by leaps and bounds. The teaching resources (teachers, classrooms, teaching equipment, etc.) and the workload of English teachers are increasing. In order to effectively improve the efficiency of college English teaching, the paper proposes to apply genetic algorithms to the actual English course scheduling problem in colleges, taking into account all the various hardware and software constraints and the expected course scheduling goals, so as to provide a clear and concise solution to the course scheduling problem plan (parallel search for optimal scheduling) and the design and coding structure of each genetic operator. Furthermore, this study creates a genetic algorithm-based English social platform and examines the design aspects of dynamic teaching models and classroom activities of college English students in the context of this paper.

Natural Language Processing (NLP) is an efficient method for enhancing educational outcomes. In educational settings, implementing NLP entails starting the learning process through natural acquisition. English teaching and learning have received increased attention from the relevant education departments as an integral aspect of the new curriculum reform. The environment of English teaching and learning is undergoing extraordinary changes as a result of the constant improvement and extension of teaching level and scale, as well as the growth of Internet information technology. As a result, the current research aims to look into techniques for efficiently using AI (artificial intelligence) apps to teach and learn English from the perspective of university students. This research can measure the levels as well as effectiveness of the employment of AI applications for teaching English based on deep learning techniques. There, the NLP based language enhancement has been carried out using Character-level recurrent neural network with back Propagation neural network (Cha_RNN_BPNN) based classification. With the help of this DL (deep learning) technique, it is possible to use AI methods to assist teachers in analysing and diagnosing students' English learning behaviour, replacing teachers in part to answer students' questions in a timely manner, and automatically grading assignments during the English teaching process. Experimental analysis shows Word Perplexity, Flesch-Kincaid (F-K) Grade Level for Readability, Cosine Similarity for Semantic Coherence, gradient change of NN, validation accuracy, and training accuracy of the proposed technique.