I am a recent Computer Science PhD graduate in the School of Computer Science at the University of Manchester. My PhD has focused on researching how we can analyse
user behaviour longitudinally in an unobtrusive way. As a result, I successfully discovered trends in Web interaction
that can be used as a proxy for familiarity. The PhD was supervised by Dr. Simon Harper and Dr. Caroline Jay. Within the school, out research
group is the
Information Management Group.
This group comprises various research labs, ranging from description logics and ontologies to distributed systems. I
am part of the Interaction Analysis and Modelling Lab,
which focuses on human factors.
Before coming to Manchester, I studied Computer Science in the University of the Basque Country.
For my final thesis I developed an Android application to explore the use of mobile devices and social networking functionalities
to enhance touristic experiences.
Feel free to contact me using the following mail:
aitor.apaolaza - AT - manchester.ac.uk
This research is the result of my PhD. The purpose of the research done during my PhD is capturing low-level Web interaction
data from Web sites and applications to support longitudinal analysis. The main contribution is the analysis of this
fine-grained interaction (mouse movements, window events) in a scalable way (it supports analysis of millions of interaction
events, from thousands of users). To do this, I aggregate interaction data into micro behaviours, which consist of
low-level aggregations of interaction data. In the context of my PhD, I used this technique to determine if and how
users' Web interaction evolves, and how this evolution can be supported through a better design.
The IDInteraction project asks: can we exploit models of human behaviour to move away from direct, unambiguous user commands,
towards seamless user-device interaction? It will investigate and develop the techniques to capture 'Indicative Usage
Models (IUMs), behavioural patterns or cues that precede a particular activity, and translate these into software-based
'Indicative Usage Patterns' (IUPs), to drive interaction with an app. The project will focus on future television broadcasting,
examining the extent to which it is possible to capture IUMs from device sensor and event data, and deploy these as
IUPs to pull content (additional information or related activities) to a 'second screen companion app', which a viewer
watches on a mobile device alongside a TV programme.
I am joining the MOVING project on September. MOVING will build an innovative training platform that will enable users from
all societal sectors (companies, universities, public administration) to fundamentally improve their information literacy
by training how to choose, use and evaluate data mining methods in connection with their daily research tasks and to
become data-savvy information professionals.
ABC: Using Object Tracking to Automate Behavioural Coding
Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors in Computing Systems, CHI 2016.
Longitudinal analysis of low-level Web interaction through micro behaviours
Proceedings of the 26th ACM Conference on Hypertext & Social Media, HT 2015. ACM
Understanding users in the wild
Proceedings of the 10th International Cross-Disciplinary Conference on Web Accessibility, W4A 2013. ACM Press
Identifying emergent behaviours from longitudinal web use
Proceedings of the adjunct publication of the 26th annual ACM symposium on User interface software and technology,
UIST 2013. ACM Press
All the source code and processing pipelines developed and used in the IDinteraction project is available here.
The project UCIVIT is composed by three different modules. The capture tool provides an easy to deploy solution to capture
Web interaction data remotely. The Analysis module contains a set of scripts employed to extract micro behaviours, and
perform longitudinal analysis on them. The visualisation module provides an interactive visualisation Web application
to ease the process of processing and analysing the data.
UCIVIT Capture tool provides the means to capture
Web interaction data unobtrusively, in a scalable way. It has shown to be able to escalate to thousands of concurrent
users, over periods longer than a year.
UCIVIT Analysis processes low-level
interaction data in a scalable way. Making use of MongoDB and the MapReduce paradigm, data captured over extended periods
of time using the UCIVIT-WebIntCap tool can be processed. Micro behaviours are small comparable units of user interaction.
These units support scalable analysis of low-level interaction without disregarding its fine-grained nature.
UCIVIT Visualisation supports researchers'
analysis of longitudinal interaction data. Distributions of interaction values at different stages of interaction are
provided, as well as regression models through the use of linear mixed models.
In case you are interested, this is the latest version of my Curriculum Vitae