Psychology + Computer Science?

Summer 2016

Tushita Patel

Schulich Leader at University of Saskatchewan

From staring at beautiful prairie skies to writing sophisticated algorithms at work – this summer has been wonderful!
 
Early Summer Retreat: Toronto
 
After my final exams ended in mid-April, I went to Toronto for about 3 weeks. It was lovely to meet my relatives, friends and high school teachers. I roamed around Toronto, like a tourist in my own city, because as a teenager living in Mississauga, I never really had an opportunity to visit many tourist attractions in Toronto. My favourite place would have to be the Aga Khan Museum in East Toronto. 
 
NSERC-USRA – Human Computer Interaction Lab
 
My research work was under the supervision of Dr. Regan Mandryk, who is not only one of the finest researchers in HCI but also a wonderful person. My work for the lab contributed mainly of audio analysis. I created tools that would extract information from recorded conversations such as engagement and composure of the speakers, and also their emotional behaviour while they were speaking. Such a tool would be highly useful for HCI research where participant’s emotional experience is valued. By the end of the summer, I participated in a Research Poster Competition organized by USRA. Here is the abstract that I submitted:
 
Investigating Relational Closeness Using Auditory Analysis
 
In research, it can be hard to evaluate study participants’ thought process and emotional experience during studies. It would be beneficial for researchers to have a tool that computationally extracts information such as warmth and engagement from recorded audio conversations of the participants. This research project is an attempt to obtain information about people’s behavioural and emotional experience in verbal conversations using three types of metrics - signal metrics: using just the signals in a recorded audio, semantic metrics: transcribing the conversations and lexically calculating the effectiveness of the conversation, and pitch metrics: using the auditory features such as frequency, modulation and pace to infer emotions of the speaker. To evaluate our approach, we ran the set of metrics on multiple conversations. We found strong correlations between our computationally derived metrics and participants’ subjective evaluations of their conversations (e.g. Affection, Composure). Our findings suggest that this tool can prove to be quite valuable for research purposes that require determining the emotions of the study participants, especially in the area of social sciences. Moreover, the extraction of such information in this project can be significantly useful in areas such as speech pathology, quality management in customer service, and mental health.
 
My experience as a summer research student was full of learning. I learned a few more languages such as Python, JavaScript and C#, and I also learned how to design such tools and also work with existing, slightly-larger programs. The project was just the right amount of challenge I would have liked: it was neither too hard nor too easy for me. By the end of September, Dr. Mandryk is hoping to submit a paper which talks about the tool that I created, and we are hoping that the paper gets accepted in CHI, the most popular international conference in Human Computer Interaction.
 
Teaching Assistantship
 
Besides working as a research assistant in the HCI Lab, I also worked 8 hours a week as a Teacher Assistant(TA) for two first-year honours level courses in computer science. This experience helped me improve my public speaking skills and my problem solving skills as it involved helping students with debugging their code. I found TA-ing (as we casually call it) to be a very rewarding experience because half way though the first course, my students gave me a middle name (since I don’t have one) of “the-Wizard”! 
 
Things Other than Work
 
Although working for 48 or more hours a week pretty much took over all the summer, I made sure that I always maintained a balance and had fun. With great determination, I self-taught myself to play the ukulele and I think I’ve become good at it!
 
I promised myself to exercise at least three hours a week, which broke soon after I got sick in early July. I stayed sick for over a month. I also volunteered with the Habitat for Humanity to build houses for people in the lower income group. 
 
Towards the end of the summer, I managed to go for a road test and succeeded in my first attempt to obtain a NOVICE-A Province of Saskatchewan Driver's License. Woot! Woot!. 
 
Academic Year III Prospects
 
My third year of university is going to be very busy. I am taking three exciting computer science courses each term, with a couple of mathematics and statistics courses as well. My program offers an optional internship year after the third year, in which I am interested to participate. I will be busy applying for internship opportunities and updating my LinkedIn and Github account throughout the year, which sounds easy, but I bet it could be very time consuming, especially when there are group assignments in majority of my courses. 
 
I am going back to working for the University’s Undergraduate Research Journal(USURJ). But this time, I am also helping out with the website developing side of the journal. I have also joined Ladies Learning Code, and I will be mentoring a couple of workshops during their events this year, which is exciting. I am highly passionate about encouraging women to pursue men-centric fields such as engineering and science. I will also be continuing to participate in programming competitions and game jams (similar to a hackathon, but for designing games) when the opportunity arises. 
 
Besides school work and extra-curriculars, I am hired again to work as a TA for a new computer science course offered to very bright first year students, and I am also continuing to work with my professor in her lab. I will be helping the research team of my supervisor in working on some little projects and also writing papers. I hope that this summer’s work will result in co-authorship of a paper!
 
 
 
 
Picture: My poster for the USRA Research Poster Competition. Photo credits to Amy Kwan. I tried to make my poster look a little different than usual. It looks like a conversational chat, signifying that more and more people are moving towards computer mediated communication.