My name is Mert Ozturk and I’m currently continuing my M.Sc. in Data Science at Sabanci University. I received my undergraduate degree in Statistics from Middle East Technical University My research interests are natural language processing and sentiment analysis of social media data. Currently, I am a member of VRL Lab and working with Onur Varol about sentiment analysis of Turkish Twitter users.

Education

• M.Sc.Data Science, Sabanci University, 2022

• B.Sc. Statistics, Middle East Technical University, 2020

• Erasmus European Mobility, Otto von Guericke University Magdeburg, 2019

Projects

Wordle Hackathon: VRL Lab hackathon to analyze Wordle dataset.

• Created a Wordle Twitter dataset by streaming tweets including Wordle results.
• Analyzed player habits by retention and success rates.
• Tested various algorithms to reach efficient solutions.

Turkish BERT for Sentiment Analysis: Fine-tuned a BERT model by Twitter data for multiclass sentiment classification task.

• Created a manually labeled Turkish Twitter dataset.
• Trained various Deep NLP models with this data.
• Obtained promising results with sentiment analysis datasets from other domains.
• Github repo: merttozturrk/Deep-NLP

Restearn Customer Analysis:

• Pre-processed the customer dataset of a customer loyalty and marketing platform.
• Created new features to make customer segmentation.
• Predicted next transcations of the customers.

Research Interests

• Sentiment Analysis

• Natural Language Processing

• Computational Statistics

• Machine Learning

• Economics

• Sports Analytics