A Graph per Persona: Reasoning about Subjective Natural Language Descriptions

Abstract

Reasoning about subjective natural language descriptions such as opinions and preferences is a challenging topic which largely hasn't been solved to date. In particular, the state-of-the-art large language models (LLMs) perform disappointing in this task, show strong biases, and do not meet the interpretability requirements we often have in this kind of applications. We propose a novel approach for reasoning about subjective knowledge which integrates potential, implicit meanings and explicitly models the relational nature of the information. We apply supervised graph learning, offer explanations for the model's reasoning, and show that our model performs well across all 15 topics of OpinionQA, outperforming several prominent LLMs. Our detailed analysis further shows its unique advantages and the complementary nature it offers in comparison to LLMs.