侯文娟Hou, Wen-Juan陳信睿Chen, Xin-Rui2022-06-089999-12-312022-06-082021https://etds.lib.ntnu.edu.tw/thesis/detail/da33cedd58c76f5a00a6cdd0cabafd82/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/117319本篇論文提出一個強健語言模型加上額外特徵的系統，處理SemEval 2019RumourEval: Determining rumour veracity and support for rumours (SemEval2019 Task 7)，主要包含了兩個任務，任務A 為 使用者的立場偵測，任務B偵測謠言是真、假或未驗證， 本研究利用到了對話分支的追蹤資訊，使用強健的預訓練語言模型與詞頻特徵，加上報導其他特徵的深度學習預訓練模型，結合兩者的預測結果，做出任務A的立場驗證，其Macro F1達到62%，再透過規則模型處理任務B的消息驗證，達到 Macro F1 68%，且 RSME降到0.5983。In this paper, we propose a robust language model with external features todeal with SemEval 2019 RumourEval: Determining rumour veracity and supportfor rumours (SemEval 2019 Task 7), which mainly contains two tasks. They areTask A: User’s stance detection, and task B: detect whether the rumour is true,false or unverified. We used the tracking of the dialogue branch, a robustly pretrained language model and word frequency features concatenate a deep learningpre-trained model that reported other features. Concatenating the prediction results of the two, we reached the performance of 62% Macro F1 for task A , andthen processed the message verification of task B through a rule-based system toreach Macro F1 68% where is RMSE is reduced to 0.5983.語言模型深度學習假新聞規則模型Language ModelDeep LearningFake newsRule-based System學術論文