Forecasting directional bitcoin price returns using aspect-based sentiment analysis on online text data

Loginova, Ekaterina; Tsang, Wai Kit; Heijningen, Guus van; Kerkhove, Louis-Philippe; Benoit, Dries F.

doi:10.1007/s10994-021-06095-3

Cited by 16 publications

(7 citation statements)

References 60 publications

(66 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[13] introduced a neural net framework that is interpretable for FSA, following a hierarchical approach paired with a query-driven attention mechanism to analyse the sentiment of news texts in the financial domain. Furthermore, [14] has illustrated that the level of interpretability is significantly increased (in addition to improved prediction results) with the use of ABSA in an application on Bitcoin directional forecasting from online text.…”

Section: Related Workmentioning

confidence: 99%

FinXABSA:Explainable Finance through Aspect-Based Sentiment Analysis

Ong¹,

Heever²,

Satapathy³

et al. 2023

Preprint

View full text Add to dashboard Cite

This paper presents a novel approach for explainability in financial analysis by utilising the Pearson correlation coefficient to establish a relationship between aspect-based sentiment analysis and stock prices. The proposed methodology involves constructing an aspect list from financial news articles and analysing sentiment intensity scores for each aspect. These scores are then compared to the stock prices for the relevant companies using the Pearson coefficient to determine any significant correlations. The results indicate that the proposed approach provides a more detailed and accurate understanding of the relationship between sentiment analysis and stock prices, which can be useful for investors and financial analysts in making informed decisions. Additionally, this methodology offers a transparent and interpretable way to explain the sentiment analysis results and their impact on stock prices. Overall, the findings of this paper demonstrate the importance of explainability in financial analysis and highlight the potential benefits of utilising the Pearson coefficient for analysing aspect-based sentiment analysis and stock prices. The proposed approach offers a valuable tool for understanding the complex relationships between financial news sentiment and stock prices, providing a new perspective on the financial market and aiding in making informed investment decisions.

show abstract

Section: Related Workmentioning

confidence: 99%

FinXABSA:Explainable Finance through Aspect-Based Sentiment Analysis

Ong¹,

Heever²,

Satapathy³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Considering how widespread cryptocurrency information has become, Loginova proposed a bitcoin price direction prediction method in 2021 that combined the sentiment analysis model JST and TS-LDA [24]. They used market trading data as well as text data from Reddit, CryptoCompare and Bitcointalk.…”

Section: Related Workmentioning

confidence: 99%

A Stacking Ensemble Deep Learning Model for Bitcoin Price Prediction Using Twitter Comments on Bitcoin

Chen

et al. 2022

Mathematics

View full text Add to dashboard Cite

Cryptocurrencies can be considered as mathematical money. As the most famous cryptocurrency, the Bitcoin price forecasting model is one of the popular mathematical models in financial technology because of its large price fluctuations and complexity. This paper proposes a novel ensemble deep learning model to predict Bitcoin’s next 30 min prices by using price data, technical indicators and sentiment indexes, which integrates two kinds of neural networks, long short-term memory (LSTM) and gate recurrent unit (GRU), with stacking ensemble technique to improve the accuracy of decision. Because of the real-time updates of comments on social media, this paper uses social media texts instead of news websites as the source data of public opinion. It is processed by linguistic statistical method to form the sentiment indexes. Meanwhile, as a financial market forecasting model, the model selects the technical indicators as input as well. Real data from September 2017 to January 2021 is used to train and evaluate the model. The experimental results show that the near-real time prediction has a better performance, with a mean absolute error (MAE) 88.74% better than the daily prediction. The purpose of this work is to explain our solution and show that the ensemble method has better performance and can better help investors in making the right investment decision than other traditional models.

show abstract

“…, 2019). Furthermore, TS-LDA (Loginova et al. , 2021), the framework of multifaceted topic model (Huang et al.…”

Section: Related Workmentioning

confidence: 99%

“…Meanwhile, with the help of ACT, disaster information could be obtained from Twitter and Instagram so as to reveal (Zhang et al, 2019). Furthermore, TS-LDA (Loginova et al, 2021), the framework of multifaceted topic model (Huang et al, 2019) was used to reflect characteristics of topic evolution of social platform. For the focus of topic evolution, when it comes to public opinion detection, researcher expounded content stream of users' search, took topics of public opinion for cluster analysis (Fathi et al, 2019), and probed into users' search actions and social network public opinion topics so as to realize the detection of the content of net-mediated public sentiment (Hasan et al, 2018).…”

Section: Analysis On the Use Of Social Media In Emergency Managementmentioning

confidence: 99%

Detection and analysis of emergency topic in social media considering changing roles of stakeholders

Xue²,

Li³

et al. 2022

OIR

View full text Add to dashboard Cite

PurposeMost of the existing studies on the evolution of emergency topics in social media focused on the emergency information demand of fixed user type in emergency while ignoring the changing roles of stakeholders during the emergency. Thus in this study, a three-dimensional dynamic topic evolution model is proposed, in which fine grained division of time, dynamic identification of stakeholders in the emergency, and emergency topic evolution based on both timeline and stakeholder's type are all considered.Design/methodology/approachParticularly the relevance between the tweets posted and the topic of emergency, the influence on the social network, and the attention of emergency topic are as well taken into account to quantitatively calculate the weight and ranking of stakeholders at different stages of the emergency. To verify the proposed model, an experimental demonstration was carried out under an emergency event posted on social media.FindingsThe results show that (1) based on the three-dimensional dynamic topic evolution model, the composition and ranking of stakeholders have obvious differences at different stages; (2) the emergency information needs and the sharing behavior of stakeholders on emergency information also indicate different preferences where the topic concerns of stakeholders at different stages have a strong relationship with their weight ranking; (3) the emergency topic evolution considering both the dynamics of emergency stakeholders and emergency information demand could more accurately reflect the changing regularity of social media users' attention to information in emergency events.Originality/valueThis study is one of first to investigate the emergency topic evaluation on social media by considering the dynamic changes of various stakeholders in emergency. It could not only theoretically provide more accurate method to understand how users share and search emergency information in social media, but also practically signify an information recommendation way in social media for emergency tracking.Peer reviewThe peer review history for this article is available at: https://publons.com/publon/10.1108/OIR-02-2021-0098.

show abstract

Forecasting directional bitcoin price returns using aspect-based sentiment analysis on online text data

Cited by 16 publications

References 60 publications

FinXABSA:Explainable Finance through Aspect-Based Sentiment Analysis

FinXABSA:Explainable Finance through Aspect-Based Sentiment Analysis

A Stacking Ensemble Deep Learning Model for Bitcoin Price Prediction Using Twitter Comments on Bitcoin

Detection and analysis of emergency topic in social media considering changing roles of stakeholders

Contact Info

Product

Resources

About