http://hdl.handle.net/10379/5419 2017-10-30T00:01:04Z 2017-10-30T00:01:04Z Wood, Ian http://hdl.handle.net/10379/6907 2017-10-12T01:01:51Z 2015-01-01T00:00:00Z

A case study of collecting dynamic social data: The pro-ana twitter community Wood, Ian The study of social processes in on-line social media content is a relatively new and rapidly growing endeavour. Many social media platforms provide a public API (Application Programming Interface) which can be used for the targeted collection of data from perceived communities, however existing software for this purpose focusses on a â snapshotâ of the community and its communications, and ignores im- portant aspects of its dynamics. We present a data collection system designed to capture tweets and the dynamics of Twitter user profile and friend/follower lists, and an approach to identify a set of tags or keywords that define an on-line community. This approach and system were used to collect a data set spanning 2 years and 7 months (including 3 Christmas periods) from the â pro-anaâ (pro-anorexia) and eating disorder Twitter community.

2015-01-01T00:00:00Z Aslam, Asra Khan, Ekram Beg, M.M. Sufyan http://hdl.handle.net/10379/6886 2017-10-07T01:00:59Z 2015-08-21T00:00:00Z

Improved edge detection algorithm for brain tumor segmentation Aslam, Asra; Khan, Ekram; Beg, M.M. Sufyan Image segmentation is used to separate objects from the background, and thus it has proved to be a powerful tool in bio-medical imaging. In this paper, an Improved Edge Detection algorithm for brain-tumor segmentation is presented. It is based on Sobel edge detection. It combines the Sobel method with image dependent thresholding method, and finds different regions using closed contour algorithm. Finally tumors are extracted from the image using intensity information within the closed contours. The algorithm is implemented in C and its performance is measured objectively as well as subjectively. Simulation results show that the proposed algorithm gives superior performance over conventional segmentation methods. For comparative analysis, various parameters are used to demonstrate the superiority of proposed method over the conventional ones.

2015-08-21T00:00:00Z Muñoz, Emir Nováček, Vít Vandenbussche, Pierre-Yves http://hdl.handle.net/10379/6749 2017-08-23T01:01:08Z 2017-08-18T00:00:00Z

Facilitating prediction of adverse drug reactions by using knowledge graphs and multi-label learning models Muñoz, Emir; Nováček, Vít; Vandenbussche, Pierre-Yves Timely identification of adverse drug reactions (ADRs) is highly important in the domains of public health and pharmacology. Early discovery of potential ADRs can limit their effect on patient lives and also make drug development pipelines more robust and efficient. Reliable in silico prediction of ADRs can be helpful in this context, and thus, it has been intensely studied. Recent works achieved promising results using machine learning. The presented work focuses on machine learning methods that use drug profiles for making predictions and use features from multiple data sources. We argue that despite promising results, existing works have limitations, especially regarding flexibility in experimenting with different data sets and/or predictive models. We suggest to address these limitations by generalization of the key principles used by the state of the art. Namely, we explore effects of: (1) using knowledge graphs machine-readable interlinked representations of biomedical knowledge as a convenient uniform representation of heterogeneous data; and (2) casting ADR prediction as a multi-label ranking problem. We present a specific way of using knowledge graphs to generate different feature sets and demonstrate favourable performance of selected off-the-shelf multi-label learning models in comparison with existing works. Our experiments suggest better suitability of certain multi-label learning methods for applications where ranking is preferred. The presented approach can be easily extended to other feature sources or machine learning methods, making it flexible for experiments tuned toward specific requirements of end users. Our work also provides a clearly defined and reproducible baseline for any future related experiments.

2017-08-18T00:00:00Z Ferrarotti, Flavio Hartmann, Sven Link, Sebastian Marin, Mauricio Muñoz, Emir http://hdl.handle.net/10379/6490 2017-04-28T01:01:11Z 2013-01-01T00:00:00Z

Soft cardinality constraints on XML data: How exceptions prove the business rule Ferrarotti, Flavio; Hartmann, Sven; Link, Sebastian; Marin, Mauricio; Muñoz, Emir We introduce soft cardinality constraints which need to be satisfied on average only, and thus permit violations in a controlled manner. Starting from a highly expressive but intractable class, we establish a fragment that is maximal with respect to both expressivity and efficiency. More precisely, we characterise the associated implication problem axiomatically and develop a low-degree polynomial time decision algorithm. Any increase in expressivity of our fragment results in coNP-hardness of the implication problem. Finally, we extensively test the performance of our algorithm. The performance evaluation provides first-hand evidence that reasoning about expressive notions of soft cardinality constraints on XML data is practically efficient and scales well. Our results unleash soft cardinality constraints on real-world XML practice, where a little more semantics makes applications a lot more effective in contexts where exceptions to common rules may occur.

2013-01-01T00:00:00Z