Numeral Understanding in Financial Tweets for Fine-Grained Crowd-Based Forecasting

Abstract

Numerals that contain much information in financial documents are crucial for\nfinancial decision making. They play different roles in financial analysis\nprocesses. This paper is aimed at understanding the meanings of numerals in\nfinancial tweets for fine-grained crowd-based forecasting. We propose a\ntaxonomy that classifies the numerals in financial tweets into 7 categories,\nand further extend some of these categories into several subcategories. Neural\nnetwork-based models with word and character-level encoders are proposed for\n7-way classification and 17-way classification. We perform backtest to confirm\nthe effectiveness of the numeric opinions made by the crowd. This work is the\nfirst attempt to understand numerals in financial social media data, and we\nprovide the first comparison of fine-grained opinion of individual investors\nand analysts based on their forecast price. The numeral corpus used in our\nexperiments, called FinNum 1.0 , is available for research purposes.\n