My main task in HCC lab was to assist research that investigates social response during the Go Challenge between Sedol Lee and AI AlphaGo created by Google DeepMind. As a research method, we used chronological data analysis on relevant online news comments and 1 on 1 semi-structured interviews for both quantitative and qualitative data. I conducted interviews and assisted tagging the result for insights with the guidance of the first author who was a Ph.D student in HCC lab.
Besides, I helped a master student who analyzed data on popular new media to see the relationship between the winning of candidates and media-exposed facial expressions during the general election on April. It was assumed that the candidates with positive, happy facial expressions would be usually elected, but according to sample testing, it was the opposite.
I looked at the pre-found pictures in articles and distinguished and labeled the names of the candidates who had facial expression data. About 1,900 facial expression data for 3 weeks before the election were collected. With Emotion API provided by Microsoft's Cognitive Services, we collated the sampled image and generated json files, formatted into spreadsheets. Each of seven expressions, Anger, Contempt, Disgust, Fear, Happiness, Neutral, Sadness, Surprise were valued between 0 and 1. While labeling, I personally felt how superficial the domestic politicians seemed to be on the media. There had always been performances like putting a funny wig, taking a meal at the civic market, making a candidate, etc. Furthermore, there was a tendency that only the representatives of the parties were back in the limelight.
Since the lab’s main research theme focused on mass social data, I gained skills from the workshop that one of the lab members hold. The course encompassed the basic data structure of Python, and finally individual assignment to implement interesting data analysis. The workshop is so useful and inspiring that I further decided to study on data science. I gained necessary, basic knowledge in regard to data science for short time.
As a part of collaborative research between User Experience Lab and iSchool, University of Washington, there was a workshop pertaining to the behavioural difference between Korean college students and USA students. I participated in the workshop as one of the mass researchers. I conducted interviews with five participants about their daily picture-taking behaviour with actual records of their sharing. Based on the interview, I participated in team discussion with UW students for two days. We analysed the I provided them the word connection data based on the code that I applied in data analysis workshop.
I could acquire how to conduct research in the academia. Even after the internship, I further took Corpus Linguistics and Python data science specialization course in Coursera to acquire further knowledge pertaining to Big Data and quantitative data analysis.