A Secure Framework for Sentiment Analysis of Cryptocurrency Data using Fin BERT Model: Ensuring Integrity and Intrusion Detection

Cryptocurrency as an elective procedure of installment that acts both as a sort of cash and as a virtual bookkeeping system has persistently been of interested to examiners. It was showed up that by applying weights to the estimation scores of tweets concurring to the opinions of the individuals, the precision of the figure will increase and a vital refinement between the exactness scores was observed utilizing the Fin BERT illustrate concurring to the MAE marker. Besides, a half breed demonstrate is proposed based on the combination of highlights removed from the compositions by one of the dictionary-based substance examination models and the highlight of weighted estimation scores. It was showed up that our proposed model illustrates defeated the other models in anticipating the costs of Bitcoin concurring to the MAE pointer. Besides, our proposed illustrate based on weighted estimation scores concurring to the affect figure of the Twitterers defeated the other models inside the forecast of conclusion of the trend of Bitcoin, which appears that the increase inside the number of highlights will not persistently lead to an increase within the exactness of our desire models. Leveraging SHA-256 algorithm, data integrity is obtained by using intrusion detection mechanisms.