A Multimodal Event-Driven LSTM Model for Stock Prediction Using Online News

Li, Qing; Tan, Jinghua; Wang, Jun; Chen, Hsinchun

doi:10.1109/tkde.2020.2968894

Cited by 121 publications

(69 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This includes discrete (GARCH [11], rolling regression [71]), continuous (Markov chain [40] & stochastic volatility [2]), and neural approaches [48,55,58,68]. Contemporary approaches: Newer work categorized under Fundamental Analysis [1] based on the Efficient Market Hypothesis [60] highlight the success of multimodal data in finance [53], as they capture a wider set of affecting knowledge and their interdependencies. Recent models used textual data such as social media posts, news reports, web searches, etc.…”

Section: Related Work 21 Multimodality In Financial Forecastingmentioning

confidence: 99%

“…Although existing research has used text for volatility prediction, only a few very recent studies exploit multimodality, especially vocal cues [77]. Multimodal approaches can extract complementary information from multiple modalities to improve financial modeling [53,77]. Financial tasks, such as predicting a stock's price movement and volatility, are often strongly correlated, thus making multi-task learning a promising modeling choice for financial forecasting.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Multimodal Multi-Task Financial Risk Forecasting

Sawhney

Mathur

Mangal

et al. 2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Stock price movement and volatility prediction aim to predict stocks' future trends to help investors make sound investment decisions and model financial risk. Companies' earnings calls are a rich, underexplored source of multimodal information for financial forecasting. However, existing fintech solutions are not optimized towards harnessing the interplay between the multimodal verbal and vocal cues in earnings calls. In this work, we present a multitask solution that utilizes domain specialized textual features and audio attentive alignment for predictive financial risk and price modeling. Our method advances existing solutions in two aspects: 1) tailoring a deep multimodal text-audio attention model, 2) optimizing volatility, and price movement prediction in a multi-task ensemble formulation. Through quantitative and qualitative analyses, we show the effectiveness of our deep multimodal approach. CCS CONCEPTS • Social and professional topics → Economic impact; • Computing methodologies → Natural language processing.

show abstract

Section: Related Work 21 Multimodality In Financial Forecastingmentioning

confidence: 99%

mentioning

confidence: 99%

Multimodal Multi-Task Financial Risk Forecasting

Sawhney

Mathur

Mangal

et al. 2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

show abstract

“…The experimental data uses the CSI 300 Index (CSI 300), which is close to 3,000 companies, traded between April 29, 2017 and December 31, 2017, with an average accuracy of approximately 60%. Many subsequent studies have also shown that machine learning and deep neural networks have superior performance when applied to the trend of time series data such as stock market trends [48], [49], [50], [51], [52], [53], [54], [55].…”

Section: Literature Reviewmentioning

confidence: 99%

Exploring the effectiveness of deep neural networks with technical analysis applied to stock market prediction

Lee

Chang

Hung

et al. 2021

ComSIS

View full text Add to dashboard Cite

The sustainable development of the national economy depends on the continuous growth and growth of the capital market, and the stock market is an important factor of the capital market. The growth of the stock market can generate a huge positive force for the country's economic strength, and the steady growth of the stock market also plays a pivotal role in the overall economic pulsation and is very helpful to the country's high economic development. There are different views on whether the technical analysis of the stock market is efficient. This study aims to explore the feasibility and efficiency of using deep network and technical analysis indicators to estimate short-term price movements of stocks. The subject of this study is TWSE 0050, which is the most traded ETF in Taiwan's stock exchange, and the experimental transaction range is 2017/01 ~ 2019 Q3. A four layer Long Short-Term Memory (LSTM) model was constructed. This research uses well-known technical indicators such as the KD, RSI, BIAS, Williams% R, and MACD, combined with the opening price, closing price, daily high and low prices, etc., to predict the trend of stock prices. The results show that the combination of technical indicators and the LSTM deep network model can achieve 83.6% accuracy in the three categories of rise, fall, and flatness.

show abstract

“…Behavioral finance, on the other hand, treats investors as an emotional entity and assumes news events to affect investments. Both the efficient market hypothesis and behavioral finance agree that news events are responsible for abnormal returns [18]. Text mining is used in event studies, and it requires words to be segmented, and spaces are used to separate words.…”

Section: Literature Surveymentioning

confidence: 99%

Building a Calendar of Events Database by Analyzing Financial Spikes

et al. 2021

View full text Add to dashboard Cite

An event is a piece of news that triggers a change in stock prices. Here, an event study is undertaken to capture the effect of abnormal returns due to an event. The event can affect the stock market in the long term or short term. Event research is relevant to both the efficient market hypothesis and behavioral finance. In this study, we collected data from websites that manage financial and economic data, performed a sentiment analysis, and correlated news article data with changes in a particular company's stock prices in the stock market. Data were collected from two well-known financial news websites. We observed a correlation between stock prices and news items. An event period of one day was considered for the study. The regression equation determined the relationship between stock returns and polarity and subjectivity. Bayesian model averaging was performed to identify the effects of polarity and subjectivity on stock returns. Time-series data were decomposed into components and detrended via regression. Prominent keywords and their polarity values for a particular day were plotted. An event enhanced some stock returns while adversely affecting other stocks. We found a variation range of -400 to 200 for different company stocks for the selected period.

show abstract

A Multimodal Event-Driven LSTM Model for Stock Prediction Using Online News

Cited by 121 publications

References 46 publications

Multimodal Multi-Task Financial Risk Forecasting

Multimodal Multi-Task Financial Risk Forecasting

Exploring the effectiveness of deep neural networks with technical analysis applied to stock market prediction

Building a Calendar of Events Database by Analyzing Financial Spikes

Contact Info

Product

Resources

About