My name is Mingyou Sung. I'm an NLP researcher & ML developer in Montreal, Quebec, Canada.

Here's some info about me!

• Languages: Korean (Native), English (Fluent)
• Programming Languages & frameworks: Python, PyTorch, HuggingFace, Scikit-Learn, NLTK, Regular Expression, GATE (General Architecture for Text Engineering), Unix/Linux, Git, GCP, Latex, AWS (Sagemaker, S3, EC2)

• Master's degree of Computer Science at Concordia University
• Bachelor's degree of Computer Engineering at Hongik University

• Relation Extraction Tasks (Studies on diverse input representations and classifiers on relation extraction datasets)

  • SemEval-2010 Task 8 (Multi-Way Classification of Semantic Relations Between Pairs of Nominals)
  • TACRED (Text Analysis Conference Relation Extraction Dataset)
  • Re-TACRED (Revised Text Analysis Conference Relation Extraction Dataset)
  • Biocreative VII Track 1 (Text mining drug and chemical-protein interactions (DrugProt))

• Counterfactual Detection (CLaC at SemEval-2020 Task 5: Muli-task Stacked Bi-LSTMs)

  • SemEval-2020 Task 5 (Modelling Causal Reasoning in Language: Detecting Counterfactuals) - Subtask 2 (Detecting the span of antecedents and consequents)

• Product Category Classification & Similar Product Identification

  • Building automatic categorization system using title and description of e-commerce products cooperate with KIES (Korea Information Engineering Services) using Word2vec and TF-IDF

• Offering Wine Information and Recommendation on Android and Web Application

  • Final year project at Hongik University (Won the excellent prize and was selected as a creditable work)
  • Developing an Android and Web application that offers wine label recognition using CNN, wine recommendation via chatbot using Regex, and a function to search for various information about wines using AWS

Teaching Assistant

• Course: COMP 479 - Information Retrieval and Web Search
• Tutored undergrads to minimize the gap between the knowledge and implementation on several information retrieval techniques such as BM25, TF-IDF