Yan Xiao

TL;DR: In this paper , a pre-training language model is employed to encode user features to fully explore potential interests of users, a linguistic knowledge model is introduced to assist the encoding, a multi-scale convolution neural network is adopted to capture text features at different scales and fully utilize the text information, and the fine-grained sentiment analysis task is modeled as a sequence labeling problem to explore the sentiment polarity of user evaluation.

TL;DR: In this paper, a moderated mediation model is built to probe into the relationship between career exploration and career adaptability, discussing the mediating role of future work self-salience and the moderated role of perceived teacher support.

TL;DR: In this paper , a two-phase mixed-methods approach was adopted to develop and validate the research model, and the results found that customer participation, including information sharing, responsible behaviors, and in-role feedback, promotes older customers' digital skills and basic needs satisfaction, leading to higher appreciation of digital affordances and well-being.

TL;DR: Wang et al. as discussed by the authors proposed a comprehensive teacher's knowledge sharing behavior evaluation method based on the improved entropy-TOPSIS method, which avoids the subjectivity of traditional expert evaluation methods and other multilevel and multi-index weight determination methods and makes the evaluation results more objective, accurate, and more realistic.

TL;DR: In this article , a biform game model of partner selection for a knowledge alliance composed of core and collaborative members is presented, where the clique solution is employed to divide the alliance's overall benefits after the introduction of the variables such as innovation efficiency, knowledge input, and marginal loss.