A Review of Sentiment, Semantic and Event-Extraction-Based Approaches in Stock Forecasting

Abstract

Stock forecasting is a significant and challenging task. The recent development of web technologies has transformed the communication channel to allow the public to share information over the web such as news, social media contents, etc., thus causing exponential growth of web data. The massively available information might be the key to revealing the financial market’s unexplained variability and facilitating forecasting accuracy. However, this information is usually in unstructured natural language and consists of different inherent meanings. Although a human can easily interpret the inherent messages, it is still complicated to manually process such a massive amount of textual data due to the constraint of time, ability, energy, etc. Due to the different properties of text sources, it is crucial to understand various text processing approaches to optimize forecasting performance. This study attempted to summarize and discuss the current text-based financial forecasting approaches in the aspect of semantic-based, sentiment-based, event-extraction-based, and hybrid approaches. Afterward, the study discussed the strength and weakness of each approach, followed with their comparison and suitable application scenarios. Moreover, this study also highlighted the future research direction in text-based stock forecasting, where the overall discussion is expected to provide insightful analysis for future reference.

Survey on sentiment analysis: evolution of research methods and topics

“…Rambocas et al examined the application of sentiment analysis in marketing research from three main perspectives, including the unit of analysis, sampling design, and methods used in sentiment detection and statistical analysis (Rambocas and Pacheco 2018). Cheng et al summarized techniques based on semantic, sentiment, and event extraction, as well as hybrid methods employed in stock forecasting (Cheng et al 2022). Yue et al categorized and compared a large number of techniques and approaches in the social media domain.…”

“…They reviewed the tasks of sentiment analysis (e.g., different text granularity, opinion mining, spam review detection, and emotion detection), the application areas of sentiment analysis (e.g. market, medicine, social media, and election prediction), and different languages for sentiment analysis, such as Chinese, Spanish, and Arabic (Adak et al 2022;Al-Ayyoub et al 2019;Alamoodi et al (2021a, b); Alonso et al 2021;Angel et al 2021;Boudad et al 2018;Brito et al 2021;Cheng et al 2022;Hussain et al 2019;Kastrati et al 2021;Khattak et al 2021;Koto and Adriani 2015;Kumar and Sebastian 2012;Ligthart et al 2021;Medhat et al 2014;Nassif et al 2021;Nassirtoussi et al 2014;Oueslati et al 2020;Peng et al 2017;Pereira 2021;Rambocas and Pacheco 2018;Ravi and Ravi 2015;Schouten and Frasincar 2015;Sharma and Jain 2020;Yue et al 2019;Zhou and Ye 2020). They summarized the methods and application prospects of sentiment analysis under different contents and topics.…”

“…Acheampong et al (2021),Ain et al (2017),Alamoodi et al (2021a, b),Asghar et al (2014),Chan et al (2022),Cheng et al (2022), Feldman (2013,Habimana et al (2020),Koto and Adriani (2015),Kumar and Sebastian (2012);Medhat et al (2014),Prabha and Srikanth (2019),Ravi and Ravi (2015),Schouten and Frasincar (2015), Serrano-Guerrero et al (2021), Taboada (2016), Wang et al (2020b), Yue et al (2019), Zhang et al (2018), and Zucco et al (2020) The existing surveys analyzed different methods of sentiment analysis. Sentiment analysis methods based on lexicons, rules, part of speech, term position, statistical techniques, supervised and unsupervised machine learning methods, as well as deep learning methods like LSTM, CNN, RNN, DNN, DBN, BERT and other hybrid approaches have been analyzed and discussed.…”

“…2021;Ain et al 2017;Alamoodi et al 2021b;Asghar et al 2014;Chan et al 2022;Cheng et al 2022;Feldman 2013;Habimana et al 2020;Koto and Adriani 2015; Kumar, Akshi and Sebastian 2012;Medhat et al 2014;Prabha and Srikanth 2019;Ravi and Ravi 2015;Schouten and Frasincar 2015;Serrano-Guerrero et al 2021;Taboada 2016;Wang et al 2020b;Yue et al 2019;Zhang et al 2018; …”

An Ensemble Multi-Layered Sentiment Analysis Model (EMLSA) for Classifying the Complex Datasets

“…Finally, sentiment analysis divides the reviews into three types such as positive, negative, and neutral, based on the text data. Sentiment analysis helps e-commerce applications increase the sales of specific products [4] [5].…”