The aim of this workshop is to provide a forum where international participants share knowledge on applying NLP to the FinTech domain. Recently, analyzing documents related to finance and economics has attracted much attention in the AI community. In the financial field, FinTech is a new industry that focuses on improving financial activity with technology. Thus, in order to bridge the gap between the NLP research and the financial applications, we organize FinNLP workshop series. One of the expected accomplishments of FinNLP is to introduce insights from the financial domain to the NLP community. With the sharing of the researchers in FinNLP, the challenging problems of blending FinTech and NLP will be identified, and the future research direction will be shaped. That can broaden the scope of this interdisciplinary research area.

Agent AI is one of the important research directions after we had significant success with multimodal large language models. It leads to the chance of using AI for business analysis, and also increases the uncertainty in scenario planning. Scenario planning finds its most potent applications in fields rife with uncertainty. Long-term strategic planning, geopolitics, and nascent industries are arenas where the linearity of traditional forecasting proves inadequate. While scenario planning isn't novel, its confluence with modern technological tools like NLP brings forth exciting prospects. With the rise of generative NLP technologies, as highlighted by recent research, there's an evolving landscape where scenario planning can be further refined, automated, and diversified. NLP's capability to parse vast textual datasets, identify emerging patterns, and even generate detailed narratives makes it a formidable tool in the scenario planner's toolkit.

The modern era, with its rapid technological advancements, geopolitical flux, and ever-changing socio-economic landscapes, encapsulates the VUCA paradigm - Volatile, Uncertain, Complex, and Ambiguous. Within this framework, traditional predictive methodologies, which often rest on linear extrapolations of existing trends, fall short. Unlike predictive models that seek precision based on historical and current data, scenario planning delves into a different realm. It doesn't merely project an extrapolated future; it crafts multiple narratives, each shedding light on a potential future. Rather than trying to predict a single path forward, scenario planning embraces the multifaceted, uncertain nature of the future. It takes into account various driving forces and uncertainties, crafting stories that present alternative futures, some of which may even seem improbable. This approach isn't about predicting the right future but about being prepared for a spectrum of possibilities. 

Submission Guidelines

We invite submissions of original contributions on methods, theories, applications, and systems on artificial intelligence, machine learning, natural language processing & understanding, big data, statistical learning, data analytics, and deep learning, with a focus on knowledge discovery in the financial services domain. The scope of the workshop includes, but is not limited to, the following areas:

• Representation learning, and distributed representation learning and encoding in natural language processing for financial document

• Language modeling on financial corpora including tabular and numerical data, and multi-modal modeling; large language models (LLMs) and applications for finance

• Graph representation learning, mining learning on graph structures from financial data

• Multi-source knowledge integration and fusion, and knowledge alignment and integration from heterogeneous data

• Synthetic or genuine financial datasets and benchmarks for baseline models

• Transfer learning applications for financial data, knowledge distillation as a method for compression of pre-trained models or adaptation to financial datasets

• Search and question answering systems designed for financial corpora

• Event discovery from alternative data and impact on organization equity price

• Environmental, social, governance (ESG) event discovery, evaluation, and impact assessment

• Historical and contemporary perspectives on scenario planning

• Generative AI models in scenario creation

• Ethical considerations in automated scenario planning

• Agent AI for collaborative scenario generation

• Real-world applications and case studies of agent AI in scenario planning

The ACL Template MUST be used for your submission(s). Accepted papers proceedings will be published at ACL Anthology. 

Long Paper:  May consist of up to 8 pages of content, plus unlimited pages for references and appendix. 

Short Paper and Demo Paper: May consist of up to 4 pages of content, plus unlimited references and appendix.

Schedule

• Submission Deadline: May 8th, 2024

  • Submission System: https://easychair.org/conferences/?conf=finnlpagentscen2024

• Paper Notification: June 4th

• Camera-Ready Deadline: June 25th

Organizers

• Chung-Chi Chen, Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, Japan
• Tatsuya Ishigaki, Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, Japan
• Hiroya Takamura, Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, Japan
• Akihiko Murai, Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, Japan
• Ryoko Nishino, Japan Advanced Institute of Science and Technology, Japan
• Hen-Hsen Huang, Institute of Information Science, Academia Sinica, Taiwan
• Hsin-Hsi Chen, Department of Computer Science and Information Engineering, National Taiwan University, Taiwan

Contact

All questions about submissions should be emailed to finnlp@nlg.csie.ntu.edu.tw