Fuzzy Classification of Multi-intent Utterances

TL;DR

This paper introduces a fuzzy classification scheme for multi-intent utterances that captures language vagueness and intent ambiguity, improving intent recognition by quantifying degrees of membership rather than binary labels.

Contribution

It is the first to model and quantify the fuzzy nature of natural language utterances in intent classification, addressing data sparsity with single-intent data to handle multi-intent classification.

Findings

Fuzzy memberships are influenced by lexical overlap and data distribution.

Defuzzified memberships closely match binary intent labels.

Different fuzzy functions and similarity measures affect classification accuracy.

Abstract

Current intent classification approaches assign binary intent class memberships to natural language utterances while disregarding the inherent vagueness in language and the corresponding vagueness in intent class boundaries. In this work, we propose a scheme to address the ambiguity in single-intent as well as multi-intent natural language utterances by creating degree memberships over fuzzified intent classes. To our knowledge, this is the first work to address and quantify the impact of the fuzzy nature of natural language utterances over intent category memberships. Additionally, our approach overcomes the sparsity of multi-intent utterance data to train classification models by using a small database of single intent utterances to generate class memberships over multi-intent utterances. We evaluate our approach over two task-oriented dialog datasets, across different fuzzy membership generation techniques and approximate string similarity measures. Our results reveal the impact of lexical overlap between utterances of different intents, and the underlying data distributions, on the fuzzification of intent memberships. Moreover, we evaluate the accuracy of our approach by comparing the defuzzified memberships to their binary counterparts, across different combinations of membership functions and string similarity measures.