TLDR — Extractive question answering is an important task for providing a good user experience in many applications. The popular Retriever-Reader framework for QA using BERT can be difficult to scale as it requires the re-processing of candidate documents in the context of a question in real time. By using phrase embeddings, we can process question and context independently which drastically reduces runtime demands. On a limited experiment I found QUIP to be 4x faster than a comparable QA model […]

This article is based on the master thesis Machine Learning in Application-Based Case Management (Rabben & Oftebro, 2022), which explored the use of machine learning in case management processes to automatically detect applications that should be rejected. Refer to the thesis for more details regarding the model outputs and more comprehensive discussions.

2022/10/21 Today I try to understand how the model Bloom works and what I can do with it.

7 ideas on how to optimize your search engine with Natural Language Processing without Transformers Written by Daniel Popek and Paweł Mielniczuk.

Photo by Waldemar    

How would you rate your experience in school out of 10? I’d probably give it a solid 8. But let me tell you about a sad reality.

What was Shakespeare thinking back then? — A word cloud

How I built a simple data science subject classifier to tag The Rockefeller Foundation’s grant database. Written by Dustin Marshall, this post builds upon many of the themes presented in an interview between the Towards Data Science Editors and The Rockefeller Foundation from last year’s Community Spotlight series.