Skip to main content

Maria Liakata Projects

Current projects

IBM Faculty Award and Warwick Research Development Fund (2013- )[PI]

Emotion sensing using heterogeneous mobile phone data

We want to use textual data entered by users in their mobile phones as a proxy for their emotions, in conjunction with information about their location and phone usage, so as to understand the effect of the environment on emotional fluctuation. We also plan to calibrate our results against the Warwick-Edinburgh mental well-being scale (WEMWBS). We anticipate that the results from this study will (a) provide an effective, relatively non-invasive means for monitoring emotional fluctuation, with potential in helping personalized mental health management (b) will allow us to understand what types of interventions in cities may be beneficial in addressing environmental factors affecting people’s mood.
Joint work with the Centre for Urban Science and Progress at NYU

I am also a partner on the PHEME and COSMOS projects.

Past projects

Leverhulme Trust Early Career Fellowship (2010-2013)

The title of my fellowship was “Reasoning with Scientific papers”.
I developed methods and a system for the automatic recognition of zones designating
core scientific concepts such as Hypothesis and Method in papers and have shown how these can be used
for creating automatic summaries.

SAPIENT Automation project (JISC) (2009-2010)[PI]

Principal investigator, manager and researcher of the SAPIENT Automation project, which employed
a researcher (18 months), a software engineer (12 months), two part-time researchers (2
months each) as well as three biology and thirteen chemistry experts for development and evaluation
purposes respectively. The project involved developing a methodology for the automatic recognition
of eleven core scientific concepts (Hypothesis, Result, Motivation etc) in scientific papers
and evaluating their usefulness in the context of automatic summaries. In collaboration with
NacTeM, University of Cambridge and the Royal Society of Chemistry Publishing.

For more details on SAPIENT Automation and current work visit

ART project (JISC) (2007-2009)

As a Postdoctoral researcher on the ART project I contributed to the creation of a set of meta-data,
the core information about scientific concepts (CISP), wrote a set of guidelines to implement CISP
as an annotation scheme (CoreSC) and supervised the development of a web-based semantic
annotation tool (SAPIENT) for the identification of core scientific concepts in research papers. I
was the driving force of the project and coordinated the creation of a corpus of 265 annotated
scientific papers through collaboration with 16 domain experts.

For details on the ART project visit

Robot Scientist Project (EPSRC) (2006-2007)

Previously I worked on the Robot Scientist project, where I developed computational and statistical
methods for the automation of experiments on yeast metabolism by creating a mechanism for
automated hypotheses generation and a model for yeast growth curves from irregular data. I also
contributed to a framework for formalising, storing and inferring conclusions from the data
obtained. The work was published in Science.

For details on the Robot Scientist project visit

HiMet (Hierarchical Metabolomics) project (BBSRC) (2005-2006)

As a PDRA on the HiMet project I used data mining to classify metabolomic data from Arabidopsis