Leveraging Large Language Models for Sentiment Analysis: Multi-Modal Analysis of Decentraland's MANA Token

TL;DR

This paper explores how combining Discord community sentiment analysis with multi-modal financial data improves cryptocurrency price prediction in Decentraland's virtual economy using advanced NLP and deep learning models.

Contribution

It introduces a multi-modal LSTM model integrating sentiment, trading volume, and market cap, demonstrating improved prediction accuracy over traditional price-only models.

Findings

Multi-modal model outperforms baseline in accuracy

Community sentiment is mostly neutral with a positive skew

Sentiment signals have significant predictive value

Abstract

Decentraland, a decentralized virtual reality platform operating within the expanding Metaverse ecosystem, utilizes its native MANA token to facilitate virtual asset transactions and governance. This study investigates the integration of Discord community sentiment with multi-modal financial data to enhance cryptocurrency price prediction within virtual world economies. We address: (1) identifying sentiment patterns within Decentraland's Discord community, and (2) evaluating the impact of multi-modal features on token return forecasting. Using a BERT-based large language model for sentiment analysis, we develop two LSTM architectures: a baseline incorporating historical prices and a multi-modal variant integrating sentiment scores, trading volume, and market capitalization. Results indicate predominantly neutral community sentiment with a positive skew. The multi-modal model significantly outperforms the price-only baseline in prediction accuracy. These findings demonstrate the predictive value of community-derived signals for virtual economy forecasting and establish a foundation for future research at the intersection of immersive virtual environments, natural language processing, and cryptocurrency market analysis.