Speech recognition technology has played an indispensable role in realizing human-computer intelligent interaction. However, most of the current Chinese speech recognition systems are provided online or offline models with low accuracy and poor performance. To improve the performance of offline Chinese speech recognition, we propose a hybrid acoustic model of deep convolutional neural network, long short-term memory, and deep neural network. This model utilizes DCNN to reduce frequency variation and adds a batch normalization layer after its convolutional layer to ensure (...) the stability of data distribution, and then use LSTM to effectively solve the gradient vanishing problem. Finally, the fully connected structure of DNN is utilized to efficiently map the input features into a separable space, which is helpful for data classification. Therefore, leveraging the strengths of DCNN, LSTM, and DNN by combining them into a unified architecture can effectively improve speech recognition performance. Our model was tested on the open Chinese speech database THCHS-30 released by the Center for Speech and Language Technology of Tsinghua University, and it was concluded that the DLD model with 3 layers of LSTM and 3 layers of DNN had the best performance, reaching 13.49% of words error rate. (shrink)

Few previous studies have investigated the relationship between global perspective and willingness to communicate in English. Hence, more studies are needed to validate their correlation. Furthermore, hardly any pertaining studies have been conducted at English Medium Instruction universities. As such, the current study aimed to fill these gaps in the context of an EMI university in China, by investigating whether GP correlates with second language WTC and what factors impact the two variables. Data were collected from students via an online (...) questionnaire and follow-up interviews. The questionnaire findings confirmed a moderate positive correlation between GP and L2 WTC. The interview data unraveled that several factors influenced students’ L2 WTC, including needs and motivations driving L2 WTC, concerns constraining L2 WTC, and intercultural cognition facilitating L2 WTC. These findings suggest that: students could be more determined to practice their English if they realize the significance of the role of English in their life; teachers could foster students’ WTC by creating a non-threatening English-speaking environment and encouraging students to communicate in English in and outside the classroom; and teachers could educate students about GP and L2 WTC, which might help to expand students’ horizon and stimulate their interests in foreign cultures and global affairs, so as to facilitate the sustainable growth of their English learning. (shrink)

As one of the important 21st-century skills, collaborative problem solving (CPS) has caught much attention in the assessment area. Two initiative approaches have been created: the human-to-human and human-to-agent modes. Between the two modes, the human-to-human interaction is much closer to the real-world situation and its process stream data can reveal more detailed information about the cognitive processes. In order to measure CPS ability effectively by this mode, how to extract indicators from the data and model it is crucial, while (...) the existing methods all have some disadvantages. In the present study, we proposed a new paradigm for extracting indicators and modeling the dyad data in the human-to-human mode. Specifically, both individual and group indicators were extracted from the data stream, which served as evidence for demonstrating CPS skills. In addition, a within-item multidimensional Rasch model was used to fit the dyad data. To test the validity of the paradigm, we developed five tasks following the asymmetric mechanism and integrated them into an online testing platform. Four hundred thirty-four students from China participated in the assessment and the online platform recorded their crucial actions with time stamps. The new paradigm was used to deal with the generated process stream data. Results showed that the model fitted well. The item parameter estimates and fitting indexes were acceptable, and students were well differentiated. In general, the new paradigm of extracting indicators and modeling the dyad data is feasible for improving the human-to-human assessment of CPS. Finally, the limitations of the current study and further research directions are discussed. (shrink)

Informed by the poststructuralist theory, this study investigates the case of Ming, a Chinese professor of English, about the impacts of his first language and second language learning experience, and the changes of social contexts on his L1 and L2 identities construction. It was found that being a learner of English as a Foreign Language, Ming’s identities development aligned with the poststructuralist theory, in which it is considered dynamic, fluid and conflicting. Ming negotiated and renegotiated his identities in various social (...) contexts in China and the United States and finally gained acceptance into the L2 academic community. This study not only analyzes Ming’s experience with his language learning and identities, but also unravels that conflicts may be part of the process of identities construction, and encourages learners to be persistent and emotionally resilient, while using certain strategies to retain a stable L1 identity so that they can navigate through the negative encounters during the second language acquisition process to sustain the development of their identities and L2 abilities. (shrink)