NLP+Vis: NLP Meets Visualization

Half-day tutorial at EMNLP Conference 2023.

Overview

Natural language and visualization (Vis) are two powerful modalities of human communication. The goal of this tutorial is to push forward the agenda of tightly integrating these two modalities. To this end, the tutorial will introduce NLP+Vis with a focus on two main threads of work:

  1. NLP for Vis: How to develop and adapt state-of-the-art NLP models for solving various visualization tasks?
  2. Vis for NLP: How to leverage visualization techniques to interpret and explain complex NLP models effectively?

The tutorial will first motivate why NLP+Vis is an important area of research and provide an overview of research topics on combining NLP and Vis techniques. Then an overview of state-of-the-art deep learning models for NLP will be covered. Next, we will provide an overview of applying visualization techniques for making NLP models interpretable and explainable. In the final part, we will focus on various application tasks at the intersection of NLP and Vis. We will conclude with an interactive discussion of future challenges for NLP+Vis applications. The audience will include researchers interested in applying NLP for visualizations as well as others who focus more generally at the intersection of machine learning and visualization.

Materials

Introduction [20 mins]

  • What is NLP?
  • What is Vis?
  • Why NLP+Vis?
  • An overview of research topics on combining NLP and Vis techniques
  • An overview of the tutorial

NLP for Vis [70 mins]

  • Encoder-decoder model
  • Attention mechanism
  • Transformer architecture
  • Language modeling and LLMs (e.g., BERT, T5, GPT)
  • Multi-modal learning (image, text, tables)
  • Huggingface library (time permitting)

Coffee Break

Vis for NLP [25 mins]

  • Intro to vis for interpretability
  • Vis tools and use cases
  • Challenges and limitations

NLP + Vis Applications [50 mins]

  • Visual text analytics
  • Natural language interfaces for visualizations
  • ChartNLP (e.g., Chart question answering, Text2Chart)
  • Natural language generation for visualization
  • Automated data-driven storytelling
  • NLP for chart accessibility
  • NLP+Vis for inclusions (e.g., promote visualization Literacy)

Future Challenges [15 mins]

  • Building benchmarks for training and evaluation
  • Data annotation challenges
  • Emerging applications

Slides

Part 1: Introduction

Part 2: NLP for Vis

Part 3: Vis for NLP

Part 4: NLP + Vis Applications, Future Challenges

Organizer

Shafiq Joty
Shafiq Joty

Salesforce AI Research and Nanyang Technological University

Enamul Hoque
Enamul Hoque

York University

Jesse Vig
Jesse Vig

Salesforce AI Research

Previous Versions of the Tutorial

A related tutorial was presented at the IEEE Vis 2023