Evaluating the Semantic Profiling Abilities of LLMs for Natural Language Utterances in Data Visualization

TL;DR

This paper evaluates the semantic understanding capabilities of four large language models in interpreting natural language utterances for data visualization, highlighting their strengths and limitations in extracting data context and visual tasks.

Contribution

It provides a systematic assessment of LLMs' abilities to comprehend and extract relevant information from uncertain natural language utterances for visualization.

Findings

LLMs are sensitive to uncertainties in utterances.

They can extract relevant data context.

Struggle with inferring visualization tasks.

Abstract

Automatically generating data visualizations in response to human utterances on datasets necessitates a deep semantic understanding of the data utterance, including implicit and explicit references to data attributes, visualization tasks, and necessary data preparation steps. Natural Language Interfaces (NLIs) for data visualization have explored ways to infer such information, yet challenges persist due to inherent uncertainty in human speech. Recent advances in Large Language Models (LLMs) provide an avenue to address these challenges, but their ability to extract the relevant semantic information remains unexplored. In this study, we evaluate four publicly available LLMs (GPT-4, Gemini-Pro, Llama3, and Mixtral), investigating their ability to comprehend utterances even in the presence of uncertainty and identify the relevant data context and visual tasks. Our findings reveal that LLMs are sensitive to uncertainties in utterances. Despite this sensitivity, they are able to extract the relevant data context. However, LLMs struggle with inferring visualization tasks. Based on these results, we highlight future research directions on using LLMs for visualization generation.