Media Technology 2009
Krogstad, Terje
Utilization and analysis of Graphically Rich Web Document Layout Automation
Document layout automation is an interesting area of research in the perspective of improving the efficiency in the design production workflow. There is a great potential for production efficacy gain by automating the workflow to some extent. An implementation of a document layout generator may cover some designer functions and thereby reducing task redundancy related to a potentially repetitive workload.
Document layout automation algorithms are typically implemented with weighting metrics for the purpose of weighting generated layout in the perspective of visual quality. Different weighting metrics and algorithms have been studied and implemented in the past years, but still they fail to match the general performance of a designer in practice. Literature suggests that many of the weighting metrics, despite their advanced mathematical implementation, lack the implementation of a persistent set of preference attributes. Weighting metrics based on an extended set of preference attributes related to designer preferences may further improve the performance of a document layout automation algorithm.
The scope of this thesis is to establish facts about document layout automation algorithms performance related to document layout quality. How do existing weighting metrics and layout automation implementations perform compared to a professional designer? In addition an extended set of preference attributes will be established and tested in a document layout automation context. How do document layout automation with an improved weighting metric perform compared to that of the designer?
An additional scope of this thesis is to provide a useful utilization of document layout automation in a web perspective. The thesis includes an discussion of how layout automation can be utilized in web documents and an analysis of the performance of web document layout automation.
Løkken, Tor
Metadata challenges in learning object management
Learning objects are the core of a new approach to the learning process. Rather than the traditional ʼseveral hour chunkʼ, learning objects provide smaller, self-contained and reusable units of learning. The benefits of the learning object approach are several. Reusability is one, a single learning object may be used in multiple contexts for multiple purposes. Being self-contained, each learning object can be taken independently, or aggregated into larger collections of content. We are, however, facing several challenges related to the creation and retrieval of the learning objects. The aim of this master thesis is to provide new knowledge in the learning object area, focusing on the creation and registration of learning object metadata in a learning object management system. The motivation for this master thesis is mainly to be able to present solutions to some of the current challenges related to learning objects, which hopefully will be of value to the people designing the software for learning object management, and in the end to the teachers and students using these systems.
Rudsengen, Bjørnar
Saliency maps and gaze behavior of people with Autism Spectrum Disorders
This MSc project plan addresses some of the basic technological difficulties within the group of Autism Spectrum Disorders (ASD), influenced by untypical social communication, social interaction and repetitive behavior. Based on previous work in the field, this preliminary study will seek to find a project description aiming to achieve new knowledge about how people with ASD use technology. Children with autism spectrum disorders have almost the same relationship to media technology as other children. They usually watch the same movies, TV programmes and videos as other children, and they play the same video- and computer games. They are perhaps even stronger related to media technology, having problems with social interactions and communication. Media usage as a personal sanctuary of repetitive behavior, can actually make these children heavier multimedia users than others.
Can a standardized saliency map apply to the ASD community gaze (visual attention) patterns? What are the metrics best suited for describing the gaze patterns of the ASD group? Based on what we know about the non-typical gaze behavior of people with ASD, there is reason to believe that the models of predictions based on prominent regions will fail to match the gaze behavior of the ASD group. The metrics are developed to match basic features, but we know that there are other features that the ASD group regard as more interesting than other parts of the population. Experimental studies with a test
group and a control group can possibly reveal important differences.
