Your browser doesn't support javascript.
A Study on BNM-cBLSTM for Financial Sentiment Analysis in European Bond Markets Based on mpBC-ELMo
2022 International Conference on Data Analytics, Computing and Artificial Intelligence, ICDACAI 2022 ; : 122-126, 2022.
Article in English | Scopus | ID: covidwho-2191838
ABSTRACT
Economic and fiscal policies devised by international organizations, governments, and central banks rely significantly on economic projections, particularly during times of economic instability such as the one we have recently seen with the COVID-19 virus spreading globally. However, the accuracy of economic forecasting and now casting models remains a challenge since modern economies are prone to multiple shocks that make forecasting and now casting activities extremely difficult, both in the short and medium range. The purpose of the paper is to identify the key aspects, which must be taken into account in big data sentiment analysis to solve the problem of forecasting and now casting tasks. The work has developed an mpBC-ELMo based BNM-cBLSTM for financial sentiment analysis in European Bond markets. The proposed framework is processed based on collection of big data that are formed into a corpus. Initially, the corpus data is subjected to preprocessing, which performs tokenization, Lemmatization, URL removal, punctuations removal, Dependency Parsing, Noun Phrases, Named Entity recognition, and Coreference Resolution in order to make the data healthier and to get a potential impact for analyzing the sentiments. Thereafter, the data is converted into a vector using mpBC-ELMo, which addresses the complex characteristics of the word as well as polarity and handles stock bond correlation invariant, behavioral biases. Finally, BNM-cBLSTM analyses the sentiments and provides an accurate as well as optimized improvisation. In comparison to existing state-of-the-art methods, experimental results show that the work tends to deliver a precise sentiment analysis and avoids erroneous prediction rate. © 2022 IEEE.
Keywords

Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: 2022 International Conference on Data Analytics, Computing and Artificial Intelligence, ICDACAI 2022 Year: 2022 Document Type: Article

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: 2022 International Conference on Data Analytics, Computing and Artificial Intelligence, ICDACAI 2022 Year: 2022 Document Type: Article