10/14/15

PhD Projects on Offer

This page contains descriptions of specific PhD projects I am interested in supervising.  Check also my Research pages for a broader view of a PhD topic.

If you are interested in any of these topics or in areas relevant to my research, then please feel free to get in touch.

Life-logging and Privacy Implications

Supervisor: Dr A.I. Abdelmoty

Keywords: Location privacy, Life-logging, Personalization, Social Web,

Mobile and pervasive computing technologies have become an integral part of everyday life. The rapidly growing availability and trends of personal data collection on mobile devices raises the need for user awareness of the range of data collected and the potential inherent information that may be inferred from this data.   For example, the fitness band around your arm is constantly counting the number of steps you are making and your mobile checkins into places are indicating which places you go to most and at what times.   This form of life-logging is potentially interesting for people and can have many useful purposes.   This project will argue that there is a lot of information inherent in these data sets that the user is not necessarily aware of.  Through data mining of integrated data sets, the project will investigate the potential of this personal data and their implications on user privacy.   The focus will, in particular, be on the use of location data as an anchor for other types of personal data.   Spatio-temporal data modelling and data mining methods will be considered for representing user tracks.   The project will also consider the question of how can the user control the amount and type of information that can be inferred from their life-logs to protect their privacy.

Commonsense Reasoning with Place Data on the Linked Data Web

Supervisor: Dr A.I. Abdelmoty

Keywords: Linked and Social Data Web, Spatial and Geographic Data modelling, management and reasoning

The Web is increasingly becoming a global information space consisting not just of linked documents, but also of linked data.  This emerging Web of Data now includes diverse data sets, such as DBpedia, GeoNames, Open Street Maps and Flickr.     One of the important factors to the success of this data Web is to ensure that the data in these data sources are linked together in a consistent and homogeneous manner.  This is particularly useful for information about geographic places.  Thus Cardiff University as a place in dbpedia is linked to the record defining the same place in GeoNames and photo referring to the university on Flickr.   Facts about this place in the three resources should also not be contradicting.    

In contrast to traditional map data, location information associated with place data are seldom detailed, but some relative place information is usually expressed to describe how the place is located with respect to other places, e.g. the university is located in the centre of Cardiff.   Traditional methods in spatial databases and geographic information systems rely on accurate and detailed representation of place location and shape on maps for computation and analysis.   Such methods will not be sufficient in the context of web resources where location information is sparse and sometimes vague.    This project will investigate commonsense spatial reasoning methods and how they can be used to manipulate vague location expressions to infer missing information and to detect inconsistent data.   The project will design appropriate methods for modelling place data on the Data Web and develop methods for dealing with such inconsistency problems and for the effective integration of multiple data resources.   The project will benefit from ongoing work on the subject of spatial reasoning, geographic place modelling and information retrieval already developed in previous projects in our research group.