Design and evaluation of a video retrieval environment for exploratory searches in a learning context

• Main Author: Loke, Cliff Choon Mun
• Other Authors: Foo Shou Boon, Schubert
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Library and information science::Knowledge management
• Online Access: https://hdl.handle.net/10356/144033
• Institution: Nanyang Technological University

The proliferation of Internet has made information more accessible to many people, including self-directed learners to support their learning needs. With technology and the Internet omnipresent, students today have grown up and immersed themselves in technologies for leisure and learning. These younger information seekers appear to be comfortable with finding information on the Internet. However, studies (Foo et al., 2014; Chang et al., 2012) have shown that they have limited higher order information literacy skills such as evaluation of information. This makes the search for useful and relevant information on the Internet challenging. Videos have been used in numerous environment to support learning. Public video repositories on the Internet, such as YouTube, Vimeo, and others serve as good resources for self-directed learning (SDL). Videos, being a richer form of medium and can engage the students better as compared to textual resources. Video is complex, where text, images, and audio can all be automatically extracted and indexed. The multi-dimensional makeup of a video has highlighted the need for better ways to interact with a video retrieval system so that retrieval can be more effective, more so for younger video seekers for SDL. The aim of this research is to study the interfaces of video retrieval systems that are deemed desirable and support the video seeking behavior of post-secondary students for SDL. The berrypicking model is used to guide this research. To address the research aim, the research will comprise of three interrelated studies. Study 1 explored the video seeking behavior of these adolescents and identify the video seeking process. 14 post-secondary students were recruited to perform two exploratory video search tasks while applying the think-aloud technique. The analysis of the results collected identified themes that are related to the video seeking behavior of the post-secondary students and the challenges encountered when performing exploratory video search tasks for learning. The themes are Selection of resources, Query formulation and reformulation, Selecting videos for preview, Previewing the video, decision for search task, Supporting query formulation and Interaction with video metadata. These findings serve as inputs to Study 2, where ideas and opinions were elicited from 20 experts with experience and skills from various fields to derive design concepts for a video retrieval environment. The thematic analysis resulted in nine design concepts for video retrieval environment in an exploratory search context. Discussions with two additional information retrieval experts were conducted to cogitate the design concepts and potential features that were derived and further elicit features for the development of the video retrieval prototype. Consequently, a set of video retrieval platform features were derived and a conceptual prototype was developed to demonstrate how these features support an exploratory video search. Study 3, an empirical evaluation on the conceptual prototype through a task walkthrough and a questionnaire that collected both quantitative and qualitative data. A total of 85 post-secondary students from the polytechnics and Institute of Technical Education (ITE) were recruited. The participants performed a task walkthrough that allowed the examination of the features using the conceptual prototype. The questionnaire was designed to capture the participants’ opinions on the utility of the features and what the participants liked and dislike about the features. The evaluation findings established the features that were deemed desirable and useful by post-secondary students for an exploratory search for learning videos. While the results cannot be generalised, they nonetheless provide an important insight to participants’ perception on the features. The research contributed to the stream of studies dedicated to video retrieval and seeking behavior, interactive design and studies as well as the design of video retrieval system to support the search for learning videos. A set of key features grouped into Concept Map, Granular interaction, Metadata and Value-adding was developed to be included in the design and delivery of video retrieval to support self-directed learning. These features have updated and support the strategies put forth by the berrypicking model. A number of limitations were noted where the research was conducted on the search using laptop and video seeking behavior is likely to differ on mobile devices. The research was limited to post-secondary students who are enrolled in vocational courses. Other post-secondary students with less vocational needs may result in different search behavior and challenges.