Keishin Nishiyama is a Staff NLP Scientist at GumGum, where he has progressed through data science and NLP roles since starting as an intern. He leverages his expertise in machine learning and sequence tagging, supported by a Masters Degree from the University of Louisville, to solve complex challenges.
He has published technical articles on specialized topics, including building a Japanese Named Entity Recognition system and developing novel methods for evaluating machine learning models.
Keishin Nishiyama
Staff NLP Scientist at GumGum
Santa Monica, California, United States
Overview
Topics They Care About
Japanese NLP
Authored a detailed technical article on building a custom Japanese Named Entity Recognition (NER) system using BiLSTM models.
ML Model Evaluation
Has written about sophisticated techniques for model comparison and data quality, such as McNemar's test and creating "memorization scores" for annotation guidelines.
Applied NLP
His career demonstrates a focus on practical applications of NLP, specifically in areas like sequence tagging tasks within the ad-tech industry.
Media Appearances
Keishin has no verified media appearances
Work History
7-2024
Staff NLP Scientist at GumGum
2-2021 - 6-2024
Senior NLP Scientist at GumGum
10-2017 - 2-2021
Data Scientist at GumGum
6-2017 - 8-2017
Data Scientist Intern at GumGum
Education
2014 - 2017
Master’s Degree from University of Louisville
More Information
Social Presence :
Prographics :
Exp : 8
Location : Santa Monica, California, United States
Job Level : N/A
Designation : Staff NLP Scientist at GumGum
