| 摘要: | A web-based collaborative reading annotation system (WCRAS) allows learners to collaborate efficiently in annotating digital texts for adding valued information, share ideas by expressing different perspectives on digital texts with annotations, and create knowledge by reading digital texts with annotations. However, an excessively large number of annotations, poor-quality annotations, or redundant annotations generated in a digital text may lead to information overloading, diverge readers' focused attention on important annotations, and raise readers' cognitive load, ultimately reducing the effectiveness of reading annotations in promoting reading comprehension. Based on the reading behaviors of learners engaged in a digital text with annotations, this work develops a web-based collaborative reading annotation system with two quality annotation extraction mechanisms (WCRAS-TQAEM) that include the high-grade and master annotation extraction approaches to filter out poor or redundant annotations from a digital text with annotations in order to facilitate the reading performance of learners and reduce their cognitive load in digital reading environments. Analytical results indicate that performing digital reading with the support of high-grade annotation extraction mechanism performs significantly better in terms of reading comprehension performance gain than performing digital reading without quality annotation extraction mechanism support. Moreover, the high-grade annotation extraction mechanism can enhance the reading comprehension of learners in four question types (i.e. Recall, main idea, inference, and application). In contrast, the master annotation extraction mechanism can only improve the reading comprehension of learners in three question types (i.e. Recall, main idea, and inference), viewing all annotations can only improve the reading comprehension of learners in two question types (i.e. Recall and inference). Finally, the learners applying WCRAS without or with the support of different quality annotation extraction mechanisms for digital reading apparently do not significantly differ in cognitive load. |
