國立臺灣師範大學教育心理與輔導學系Tseng, H. C.Chang, T. H.Sung, Y. T.2014-12-022014-12-022012-11-15http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/40801Readability refers to the extent to which a text can be understood. Reading materials of high readability can facilitate reading comprehension, learning, and information retention. With the development of Cloud computing, it has become important to assess the readability of online texts. However, previous readability research focuses only on the accuracy of text classification but fails to comprehensively take into account the formulae accuracy, time complexity, and feature diversity. Currently readability formulae used in evaluating online documents are time-consuming and low in accuracy. Therefore, to enhance the online utilization of readability formulae, we propose a standardized framework for evaluating the feasibility of online readability formulae, based on the three major factors: (a) user waiting time, (b) text classification accuracy, and (c) feature diversity. We also verified the feasibility of this framework by comparing several readability models under such framework.