dzFinNlp at AraFinNLP: Improving Intent Detection in Financial Conversational Agents

TL;DR

This paper describes the dzFinNlp team's approach to improving intent detection in financial conversational agents using various machine learning and deep learning models, achieving high accuracy on the ArBanking77 dataset.

Contribution

The paper introduces a combination of traditional, deep learning, and transformer models for intent detection in financial dialogue systems, with optimized feature configurations.

Findings

Best model achieved 93.02% micro F1-score on development set

Achieved 67.21% micro F1-score on test set

Demonstrated effectiveness of transformer-based models in financial intent detection

Abstract

In this paper, we present our dzFinNlp team's contribution for intent detection in financial conversational agents, as part of the AraFinNLP shared task. We experimented with various models and feature configurations, including traditional machine learning methods like LinearSVC with TF-IDF, as well as deep learning models like Long Short-Term Memory (LSTM). Additionally, we explored the use of transformer-based models for this task. Our experiments show promising results, with our best model achieving a micro F1-score of 93.02% and 67.21% on the ArBanking77 dataset, in the development and test sets, respectively.