2021/06/11 by Christian Otto, Otto, Christian, Ran Yu +17 · 1 citation
Computer Science · Psychology · #Educational Strategies and Epistemologies #FOS: Computer and information sciences #Information Retrieval (cs.IR) #Information Retrieval and Search Behavior #Psychological and Educational Research Studies
paper · pdf · doi:10.48550/arxiv.2106.06244
openalex publication_date 2021/06/11 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
In informal learning scenarios the popularity of multimedia content, such as\nvideo tutorials or lectures, has significantly increased. Yet, the users'\ninteractions, navigation behavior, and consequently learning outcome, have not\nbeen researched extensively. Related work in this field, also called search as\nlearning, has focused on behavioral or text resource features to predict\nlearning outcome and knowledge gain. In this paper, we investigate whether we\ncan exploit features representing multimedia resource consumption to predict of\nknowledge gain (KG) during Web search from in-session data, that is without\nprior knowledge about the learner. For this purpose, we suggest a set of\nmultimedia features related to image and video consumption. Our feature\nextraction is evaluated in a lab study with 113 participants where we collected\ndata for a given search as learning task on the formation of thunderstorms and\nlightning. We automatically analyze the monitored log data and utilize\nstate-of-the-art computer vision methods to extract features about the seen\nmultimedia resources. Experimental results demonstrate that multimedia features\ncan improve KG prediction. Finally, we provide an analysis on feature\nimportance (text and multimedia) for KG prediction.\n