黃啟祐Huang, Chi-Yo王星諺Wang, Sing-Yan2023-12-089999-12-312023-12-082023https://etds.lib.ntnu.edu.tw/thesis/detail/a07bc068f6693b37c81fc4f705c59a0f/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/120765近年來，由於無線通訊技術與網路發展快速，已經逐漸可支援萬物互聯，自動駕駛亦不例外。目前，各主要車廠均積極投入自動駕駛技術開發，並已經能支持如自動跟車(Adaptive Cruise Control，ACC)或自動煞停(Autonomous Emergency Braking，AEB)等第二級自動駕駛的功能，少數車廠之自駕車，已經可以達到第三級(Level 3，L3)之有條件自動駕駛。雖然自駕車市場占有率日增，然而，對影響一般消費者接受自駕車因素之相關研究仍然有限，特別針對社群網路探勘者甚少。唯近年來，社群網路已經廣為消費大眾利用，發表購買意圖或使用意見，因此，本研究擬跨越研究缺口，探勘主流社群網站，瞭解自駕車之消費行為。本研究以世界著名社交網站Twitter為樣本，利用社交媒體探勘和偏小平方結構方程模型驗證影響自動駕駛汽車接受度的關鍵因素。本研究結果顯示，兼容性(compatibility)與易用性(ease of use)對認知態度(perceived attitude)之關聯顯著，可試用性(trialability)與感知有用性(perceived usefulness)對於持續使用意願(intention of continuous attitude)有顯著關聯，此外，易用性(ease of use)與相容性(compatibility)之相關係數最高，為最重要因素。本研究結果可作為自駕車廠商訂定行銷策略之依據，發展完善之研究架構，亦可作為未來探勘各類新產品消費行為之用。In recent years, due to the rapid development of wireless communication technology and the Internet, the concept of the Internet of Things has gradually become feasible, and autonomous driving is no exception. Currently, major automotive manufacturers are actively engaged in the development of autonomous driving technology, and functionalities like Adaptive Cruise Control (ACC) or Autonomous Emergency Braking (AEB) that support Level 2 autonomous driving are already in place. A few car manufacturers have even achieved Level 3 conditional autonomous driving. Despite the increasing market share of autonomous vehicles, research on factors influencing the acceptance of autonomous vehicles by general consumers remains limited, especially regarding social media miners. However, in recent years, social media has been widely utilized by the general public to express purchase intentions or opinions on usage. Therefore, this study aims to bridge this research gap by exploring mainstream social media platforms to understand consumer behavior toward autonomous vehicles.Using the renowned global social networking platform Twitter as the sample, this study employs text mining and partial least squares structural equation modeling (PLS-SEM) to verify the key factors affecting the acceptance of autonomous vehicles. The study results reveal that the correlations between compatibility and ease of use to perceived attitude, trialability and perceived usefulness to intention of continuous attitude, as well as ease of use and compatibility are significant. Additionally, the correlation coefficient between ease of use and compatibility is the highest, indicating its paramount importance. These research findings can serve as the foundation for autonomous vehicle manufacturers in devising marketing strategies and developing refined research frameworks. Furthermore, these findings can be used as a basis for future exploration of consumer behaviors towards various types of new products.自駕車社群媒體探勘文字探勘偏最小平方結構方程式隱含狄利克雷分布演算法Self-Driving CarSocial Media MiningText MiningPartial Least Square- Structural Equation Modelling (PLS-SEM)Latent Dirichlet Allocation (LDA)etd