As the years pass, the uses and applications of NLP continue to grow firmly establishing it as one of the essential topics for data scientists. Feeling the need to brush up on your NLP knowledge? ODSC Europe will feature several expert-led sessions on the topic. Check out a few of them below.
Advanced NLP: Deep Learning and Transfer Learning for Natural Language Processing
Dipanjan (DJ) Sarkar｜Lead Data Science Consultant at Schaffhausen Institute of Technology Academy｜Google Developer Expert – ML
A highly knowledgeable, experienced, and recognized data scientist, in 2020 DJ Sarkar was named one of India’s top ten data scientists, was included in the list of 40 under 40 Data Scientists in 2021, and was recognized as a Top 50 AI Thought Leader by Global AI Hub.
DJ Sakar’s interactive, hands-on session will guide you through rigorous examples to help you gain mastery of the tools and techniques, such as deep transfer learning, required to apply NLP to real-world problems. In particular, DJ Sakar will focus on applications like language translation, classification, Q&A systems, and more.
Leonardo De Marchi｜Head of Data Science and Analytics ｜MagicLab
During an extensive career, Leonardo De Marchi has held a diversity of roles, including as the lead instructor at ideal.io, a data scientist in the professional sports industry, the author of “Hands-On Deep Learning” and a contractor.
In this half-day training session, Leonardo De Marchi will equip you with the NLP fundamentals, including pre-processing techniques,tf-idf, embeddings, and more through both theory and hands-on coding examples in Python. By the end of the session, you should have the tools you need to understand the most recent NLP topics and solve real-world problems.
Evaluating Generalisation in Natural Language Processing Models
Dieuwke Hupkes, PhD｜Research Scientist, Scientific Manager of ELLIS Amsterdam｜Facebook AI Research, European Lab for Learning and Intelligent Systems (ELLIS)
Dieuwke Hupkes’ specialty lies in researching how NLP models can be developed to show human-like behavior. She is interested in understanding the current state of the technology and where it can be improved.
Recent research has shown that models that receive positive evaluations on random within-distribution train/test splits can still have significant generalization failures. In this session, Dr. Hupkes will address the different types of generalization and the difficulties of evaluating them with a focus on NLP models.
Social Biases in Text Representations and their Mitigation
Danushka Bollegala, PhD｜Professor in the Department of Computer Science｜University of Liverpool
Dr. Bollegala is a highly recognized researcher and data scientist who has been awarded with the IEEE Young Author Award and best paper awards at GECCO and PRICAI.
The development of BERT, GPT-3, RoBERTa, and other language models have led to an increase of NLP applications. However, they also contain social biases that make them unideal for use in real-world applications. This session will focus on the methods the NLP community has developed to identify these biases and how these methods can be used to mitigate them.
Ethics Aren’t Just for AI Algorithms
Matt Beale｜Senior Solutions Consultant｜Cloudfactory
Matt Beale’s current work is focused on solving the data challenges in AI and machine learning projects in the fields of FinTech, Autonomous Vehicles, and Green Energy. He is particularly interested, however, in the ethical issues inherent in the development of these technologies.
AI’s prevalence in today’s society means it’s essential to ensure that ethics is front and center throughout the full lifecycle of your AI projects. This talk will provide you with four practical steps you can use to ensure your team is taking an ethics-centric approach when developing your AI models.
Sentiment Analysis Tricks with Keras, spaCy, and Transformers
Duygu Altinok, PhD｜Senior NLP Engineer｜Deepgram
Dr. Altinock is the author of the bestseller Mastering spaCy and is currently interested in working in dialogue management.
This workshop will teach you all the tips and tricks you need to successfully deploy a sentiment analysis model, an increasingly important skill for data scientists working with NLP in the industry. During this session, you will become familiar with using Keras, spaCy, and Transformers for sentiment analysis.
Applying Interactive Weak Supervision to NLP Tasks
Shayan Mohanty｜Co-Founder & CEO｜Watchful
During Shayan Mohanty’s career, he has co-founded a company, Watchful, led data engineering teams at several companies, given talks on a diversity of topics, and has been a Guest Scientist at Los Alamos National Laboratory.
In this session, you will learn how to effectively use weak supervision for a wider range of tasks, despite its inherent limitations and dependency on the number of available sources. This tutorial will show you how to augment a weakly supervised system with generative techniques.
Register for ODSC Europe 2022 Here
For a full list of the ODSC Europe’s virtual and in-person sessions, check out our conference schedule here. And be sure to register for your ODSC Europe pass soon, as our limited-time offer of 30% off ends Friday!