Joseph Marvin Imperial

Kumusta! I'm Joseph. I'm a professional lecturer at the Department of Computer Science at the National University, Philippines, and a senior NLP researcher at the Human Language Technology (HLT) Lab. I conduct research on the intersection of machine learning and NLP, specifically in the areas of text complexity, story generation, and social media. I'm most known for my contributions to machine learning-based Filipino text readability assessment which you can find the resources on my Github account or the list of published papers on my Google Scholar page.

On this site, you'll learn more about my expertise and past works. Send me an email if you want to collaborate on a project or a research paper in line with my interests.

Updates

June 3, 2022 - I have been featured by National University Philippines for this month's Nationalian Spotlight. Read the article here.

May 22, 2022 - I have been accepted to the prestigious Oxford Machine Learning Summer School. See website here.

May 13, 2022 - My student group at NU HLT Lab has two papers on readability assessment of Philippine languages (Filipino and Cebuano) accepted at AIED 2022 and BEA co-located at NAACL 2022. See Papers section.

April 5, 2022 - We release the first-ever baseline Machine Learning model for readability assessment in the Cebuano Language. Paper and code can be found here. Special thanks to my undergrad advisees Michael, Ranz, Lloyd, and Mohammed.

February 1, 2022 - My unified, IPA-based model placed 3rd overall for the CMCL Shared Task on Multilingual and Cross-lingual Prediction of Human Reading Behavior. Paper and code here.

Sep 30, 2021 - I have two papers accepted at PACLIC 2021 on interpreting readability assessment models and science mapping of NLP research in the Philippines. See Papers section.

Aug 17, 2021 - I am a recipient of the inaugural Google AI and TensorFlow Faculty Award to support our Machine Learning specialization at National University, Philippines. More information on this award can be found here.