Spring 2017

The department runs both a research seminar series with external speakers, as well as a more informal research colloqium that aims to bring together researchers across Edge Hill University.

Research Seminar

Events, Meta-Knowledge and Extremism – The Adventures of a Text Miner

Dr Raheel Nawaz, School of Computing, Mathematics and Digital Technologies, Manchester Metropolitan University
Thursday 30th of March 2017, 12.30pm, TechHub Lecture Theatre

Part-1: An introduction to event extraction followed by an introduction to meta-knowledge – what is a textual event? How can it de described and extracted? Why is this useful? What is meta-knowledge? Why is it useful?

Part-2: Brief introduction to mining extremist/radical/hate speech. How to identify extremist/radical/hate speech? How to characterise it? How to extract it? Why is it useful?

Research Colloquium

Can computers understand text? Text mining methods for machine translation, word sense disambiguation and document classification.

Georgios Kontonatsios, Department of Computer Science
Wednesday 25th of January 2017, 1pm, TechHub Meeting Room

Text mining is an emerging field of artificial intelligence that deals with computer processing and understanding of natural language. In this talk, we will cover intelligent text mining approaches that employ rule-based and machine learning methods to make sense of natural language encoded in text. Specifically, we will discuss machine translation methods that are able to translate terms across languages, word sense disambiguation techniques that identify the correct meaning of ambiguous words and text classification methods that determine the topic a given document. Finally, deep learning representation techniques will be introduced that have the potential to further improve the performance of text mining algorithms.

Detecting and Recognising Objects: My Practice in Imagery

Dr Huaizhong Zhang
Wednesday 8th22nd of February 2017, 2pm, TechHub Meeting Room

Object detection, tracking and recognition in images are key problems in computer vision. A variety of techniques have been involved into this scientific domain, particularly variational and stochastic aspects. Referring to my previous studies and activities, this talk provides the audience with a balanced treatment between the theory and practice of relevant methods in the area.
Emphasis on the following points:

  • Statistical based methods for boundary finding
  • Graph based ROI detection in medical images
  • Level sets and PDE for image segmentation
  • Deformable modelling for object detection and reconstruction
  • Deep learning based object recognition and tracking

Data analytics for understanding population movement

Isa Inuwa Dutse, Department of Computer Science
Wednesday 8th of March 2017, 2pm, TechHub Meeting Room

Social media is now an indispensable space that support complex networks and interactions of various users via popular public sites such as Twitter and Facebook. Focus of this talk is on the use of social media data toward understanding aspects of humans livea as discuss in social media space. Consequently, the focus will be on how to explore the social media data pertaining population movement, appropriate method to use and the applicability of text mining related techniques will be presented.

Understanding the impact of introducing climate modelling to a community of users

Pradeep Hewage, Department of Computer Science
Wednesday 29th of March 2017, 2pm, TechHub Meeting Room

Non-predictive or inaccurate weather forecasting can severely impact the farmer’s ability to engage in activities like ploughing, fertilising, cultivating, and others; resulting in a direct impact on the resources that a farmer requires to carry out these operations. This research aims to configure a numerical weather prediction system called Weather and Research Forecasting (WRF) model based on other external parameters and the historical data to obtain an operational, costumes, and more accurate weather prediction system for the farmers. This research requires the combination of the predictive data from WRF output, historical data, physics data from sensor devices which are located across field systems, and farm data to get an intelligent and reactive output for an award-winning precision farming service provider. These data will be collated and used to train a neural network to identify emergent patterns for the purposes of making future predictions, and thereby decisions, informed by historical events.

Last updated on Last updated on Was this page helpful? Yes NoThanks for your feedback!Please tell us more: