keywords:
Python, Research, Tools, Data
Context:
Research at CSSL
code and publications:
Most of the code I wrote is on the lab machines, but these two are available on GitHub
Papers contributed to
Modelling Performance on Cognitive Tests using LSTMs and Skip-Thought Vectors
And a few that are still being worked on
Description:
I worked with the Computational Social Sciences Lab at the University of Southern California for 3 years.
Major projects I have contributed to include:
studying hate speech in social media using data from the alt-right racist social network Gab
classification of moral sentiment and topic in tweets using data from the Baltimore protests
network analyses of Gab and Twitter followings based on the above two points
predicting cognitive performance based on media consumption
My primary role was tool creation, including:
data acquisition tools
for generating corpora , samples, and social network graphs
as well as database management tools to store and organize it all
combining workflows into general-purpose tools
interfaces for these tools
keywords:
Python, Research, Big Data, Massively Parallel Programming, Sentiment Analysis, Hadoop, Spark, MongoDB
Context:
Internships at China Mobile Hong Kong and Financial Data Technologies
code and publications:
I cannot provide any code or documents on the basis of confidentiality agreements with my employers.
Description:
I have worked two internships in data science with China Mobile Hong Kong and FDTAI, the AI research division of Financial Data Technologies Ltd.
At China Mobile, I wrote scripts to process hourly mobile usage data from their 1.5 million customers using Hadoop, PySpark, and MongoDB. I also initiated and laid the groundwork for a new large-scale KPI analysis project.
At FDTAI, I researched and developed stock market analysis and prediction techniques. Most notably, I developed an experimental system which scraped financial news websites for articles and performed sentiment analysis to predict future changes to the market.