PhD Research in Media and Arts Technology
My PhD research in Media and Arts Technology focuses on people’s experience with text-based interactive systems. I investigate how familiarity with technologies affect people’s perception by applying quantitative and qualitative research methodologies.
This research explores how people interact with multi-modal text in digital artworks by defining, designing and evaluating multi-modal reading experiences. First, we looked at the literature on multi-modal reading and created a taxonomy to define current interaction techniques required from readers. Then, we developed a series of studies, both in lab and in public to design and test people’s reaction to familiar and novel interaction techniques while reading multi-modal text. Our first study looks at the effects of different forms of presentation on reading experience. Our results show that readers prefer easy-to-read and familiar complexity levels for reading ease, and that they prefer challenging complexity levels when developing novel reading strategies. Secondly, our field study looks at evaluating the design of an interactive system in public. The questionnaire responses suggest that the installation was attractive (attractiveness), is easy to get familiar with (perspicuity), excites visitors (stimulation), and innovative (novelty). Our second lab study looks at the effects of familiar vs. novel input modalities on reading experience. The preliminary results suggest that readers show similar comprehension across conditions, while they prefer reading in a novel setting by using an unfamiliar input device. Our third study looks at the effects of different levels of interaction complexity on reading ease.
MSc Research in Cognitive Science & Language Technology
I completed my master studies in the Erasmus Mundus European Masters Program in Language and Communication Technologies (LCT), and my research was about combining metaphor comprehension models with knowledge mismatch.
In this thesis, it is aimed to answer whether the processing of a metaphor is different when the addressee is naive to the context of the metaphor use compared to when the addressee knows about the context of the metaphor. Therefore, an eye-tracking experiment, which allows dissociating between early and later processing stages of metaphor comprehension, was conducted to show whether different contextual factors influence the time it takes to process novel metaphors. The analyses of regression path durations suggest that supportive context facilitates processing of the metaphorical expressions but only when the addressee is aware of such context. The results reveal the importance of perspective taking in real-time metaphor comprehension.
The abstract and the thesis can be found in ︎︎︎ LCT-theses archive
Linguistic Data Annotation
I completed a 5-month work placement at NLP unit in Bruno Kessler Foundation (FBK) in Povo, Italy. I worked for the ‘Creative Language Recognition and Generation for Advertising’ project supervised by Carlo Strapparava.
My roles included: (1) researching existing datasets for figurative language, and adapting existing datasets for current research; (2) identifying sources of online ads; (3) adding >100 online ads to the database; (4) annotating online adds in the following way; determine if each slogan has any rhetorical figures based on ’Taxonomy of Rhetorical Figures’, determine if each slogan has positive, negative or neutral emotion, determine if there is a use of negation, note down confidence level. At the last phase of the project, I also worked with other researchers from FBK to (5) agree/disagree on annotation.