الفهرس | Only 14 pages are availabe for public view |
Abstract The thesis addresses how to detect psychological human traits from text and introduces a pragmatic solution by proposing an efficient and explainable new model for personality classification tasks based on transfer learning and fusion techniques. We do not rely on any hand-crafted features or affect related lexicons in the text representation step for all the models. Moreover, we examined the state-of-the-art deep learning-based models in NLP and compare the performance for these models to determine the best model under the personality detection task and find the traits which have obvious cues that are most easily predicted in the text data |