No abstract

Association rules are a class of important regularities in databases. They are found to be very useful in practical applications. However, the number of association rules discovered in a database can be huge, thus making manual inspection and analysis of the rules difficult. In this paper, we propose a new framework to allow the user to explore the discovered rules to identify those interesting ones. This framework has two components, an interestingness analysis component, and a visualization component. The interestingness analysis component analyzes and organizes the discovered rules according to various interestingness criteria with respect to the user's existing knowledge. The visualization component enables the user to visually explore those potentially interesting rules. The key strength of the visualization component is that from a single screen, the user is able to obtain a global and yet detailed picture of various interesting aspects of the discovered rules. Enhanced with color effects, the user can easily and quickly focus his/her attention on the more interesting/useful rules.

Subjective quality assessment is an essential component of modern image and video processing, both for the validation of objective metrics and for the comparison of coding methods. However, the standard procedures used to collect data can be prohibitively time-consuming. One way of increasing the efficiency of data collection is to reduce the duration of test sequences from the 10 second length currently used in most subjective video quality assessment experiments. Here, we explore the impact of reducing sequence length upon perceptual accuracy when identifying compression artefacts. A group of four reference sequences, together with five levels of distortion, are used to compare the subjective ratings of viewers watching videos between 1.5 and 10 seconds long. We identify a smooth function indicating that accuracy increases linearly as the length of the sequences increases from 1.5 seconds to 7 seconds. The accuracy of observers viewing 1.5 second sequences was significantly inferior to those viewing sequences of 5 seconds, 7 seconds and 10 seconds. We argue that sequences between 5 seconds and 10 seconds produce satisfactory levels of accuracy but the practical benefits of acquiring more data lead us to recommend the use of 5 second sequences for future video quality assessment studies that use the DSCQS methodology.

Unexpected rules are interesting because they are either previously unknown or deviate from what prior user knowledge would suggest. In this paper, we study three important issues that have been previously ignored in mining unexpected rules. First, the unexpectedness of a rule depends on how the user prefers to apply the prior knowledge to a given scenario, in addition to the knowledge itself. Second, the prior knowledge should be considered right from the start to focus the search on unexpected rules. Third, the unexpectedness of a rule depends on what other rules the user has seen so far. Thus, only rules that remain unexpected given what the user has seen should be considered interesting. We develop an approach that addresses all three problems above and evaluate it by means of experiments focusing on finding interesting rules.

No abstract