Traffic congestion is a condition of a segment in the road network where the traffic demand is greater than the available road capacity. The detection of unusual traffic patterns including congestions is a significant research problem in the data mining and knowledge discovery community. However, to the best of our knowledge, the discovery of propagations, or causal interactions among detected traffic congestions has not been appropriately investigated before. In this research, we introduce algorithms which construct causality trees from congestions and estimate their propagation probabilities based on temporal and spatial information of the congestions. Frequent sub-structures of these causality trees reveal not only recurring interactions among spatio-temporal congestions, but potential bottlenecks or flaws in the designs of existing traffic networks. Our algorithms have been validated by experiments on a travel time data set recorded from an urban road network.

Text mining in clinical domain is usually more difficult than general domains (e.g. newswire reports and scientific literature) because of the high level of noise in both the corpus and training data for machine learning (ML). A large number of unknown word, non-word and poor grammatical sentences made up the noise in the clinical corpus. Unknown words are usually complex medical vocabularies, misspellings, acronyms and abbreviations where unknown non-words are generally the clinical patterns including scores and measures. This noise produces obstacles in the initial lexical processing step as well as subsequent semantic analysis. Furthermore, the labelled data used to build ML models is very costly to obtain because it requires intensive clinical knowledge from the annotators. And even created by experts, the training examples usually contain errors and inconsistencies due to the variations in human annotators' attentiveness. Clinical domain also suffers from the nature of the imbalanced data distribution problem. These kinds of noise are very popular and potentially affect the overall information extraction performance but they were not carefully investigated in most presented health informatics systems. This paper introduces a general clinical data mining architecture which is potential of addressing all of these challenges using: automatic proof-reading process, trainable finite state pattern recogniser, iterative model development and active learning. The reportability classifier based on this architecture achieved 98.25% sensitivity and 96.14% specificity on an Australian cancer registry's held-out test set and up to 92% of training data provided for supervised ML was saved by active learning.