dzFinNlp at AraFinNLP: Improving Intent Detection in Financial Conversational Agents

Memory (LSTM) networks (Firdaus et al., 2021) The Arabic Financial NLP (AraFinNLP) shared and their bidirectional variants (BiLSTM) (Sreelakshmi task highlights the increasing importance of advanced et al., 2018), has provided more nuanced Natural Language Processing (NLP) tools understanding by capturing the sequential nature tailored for the financial sector in the Arab world. of text. More recently, transformer-based models, This initiative is particularly timely given the substantial like BERT (Alshahrani et al., 2022), have set new growth of Middle Eastern stock markets, benchmarks in NLP by leveraging self-attention driven by diverse sectors across the region. This mechanisms to understand contextual relationships economic expansion underscores the need for sophisticated within text, making them particularly effective for financial NLP systems capable of handling complex tasks like intent detection across varied the unique linguistic and cultural nuances of dialects.