30210

ارائه مدل پيشبيني قيمت ارزهاي ديجيتال مبتني بر تحليل احساسات و الگوريتمهاي يادگيري ماشين

كارشناسي ارشد

مهندسي فناوري اطلاعات گرايش تجارت الكترونيكي

1399

1402/06/28

دكتر بابك اميري

مهندسي صنايع

ارز ديجيتالي به عنوان يك روش جايگزين براي پرداخت كه با استفاده از الگوريتم‌هاي رمزگذاري ايجاد مي‌شود، هم به عنوان نوعي ارز و هم به عنوان يك سيستم حسابداري مجازي عمل مي‌كند. سرمايه گذاري در اين بازار همواره با خطراتي ناشي از عدم آگاهي كافي افراد در خصوص افزايش يا كاهش قيمت ارزهاي ديجيتالي مواجه شده است. در سال‌هاي اخير مدل‌هاي ‌‌‌پيش‌بيني قيمت ارز‌هاي ديجيتالي براي كاهش ريسك سرمايه‌گذاري در اين بازار پيشنهاد شده است. اين موضوع سبب شده است تا افراد با آگاهي بيشتري به خريد يا فروش ارزهاي ديجيتالي اقدام كنند. از آنجا كه احساسات عمو‌مي ‌يك جامعه مي‌تواند با تاثيرگذاري بر ميزان تقاضا براي خريد يا فروش ارزهاي ديجيتالي، بر قيمت آن‌ها اثرگذار باشد و شبكه‌هاي اجتماعي محلي براي انعكاس نظرات عمو‌مي‌ يك جامعه در ارتباط با موضوعات مختلف است، در اين پژوهش، مدلي بر پايه يادگيري ماشين و مبتني بر تحليل احساسات كاربران شبكه هاي اجتماعي پيشنهاد شده است كه در آن با استفاده از مدل ‌هاي يادگيري ماشين بر اساس امتيازات احساسات متون انتشار يافته توسط افراد در شبكه هاي اجتماعي ، به پيش بيني قيمت ارزهاي ديجيتالي مانند بيت‌كوين، اتريوم، اياس، كاردانو و ريپل پرداخته شده است. نتيجه حاصل از اين پژوهش نشان مي‌دهد كه دقت پيش بيني قيمت ارزهاي ديجيتالي با لحاظ كردن عواملي چون ميزان اثرگذاري افراد و شناسايي نوع احساسات موجود در متون منتشرشده در شبكه هاي اجتماعي افزايش خواهديافت.

1402/09/27

Cryptocurrency Price Prediction Model Based on Sentiment Analysis and Machine Learning Algorithms

9/18/2024 12:00:00 AM

Cryptocurrency is a digital currency and an alternative method of payment that is created using encryption algorithms, acting as both a form of currency and a virtual accounting system. Investing in this market has always faced risks due to the lack of sufficient knowledge of people regarding the increase or decrease in the price of digital currencies. In recent years, cryptocurrencies price prediction models have been proposed to reduce the risk of investing in this market. This issue has caused people to buy or sell cryptocurrencies with more awareness. Since the public sentiments of a society can influence the price of cryptocurrencies by affecting the demand for buying or selling, and social networks are used to reflect the public opinions of a society in relation to various issues, in this research, a machine learning model based on sentiment analysis of tweets has been proposed to predict the future prices of various cryptocurrencies like Bitcoin, Ethereum, EOS, Cardano and Ripple. The result of this research showed that the accuracy of the prediction model will increase by considering factors such as the influence of people and the type of emotions in the published texts in social media. It was shown that by applying weights to the sentiment scores of tweets according to the influence factor of the individuals on social media, the accuracy of the prediction will increase and a significant difference between the accuracy scores was observed by the use of the LSTM model according to the MAPE indicator. Also, a hybrid model is proposed based on the combination of features extracted from the texts by one of the dictionary-based text analysis models and the feature of weighted sentiment scores. It was shown that our proposed hybrid model outperformed the other models in predicting the prices of Ethereum, EOS, and Cardano. Also, our proposed model based on weighted sentiment scores according to the influence factor of the Twitterers outperformed the other models in the prediction of the future prices of Bitcoin and Ripple, which indicates that the increase in the number of features will not always lead to an increase in the accuracy of our prediction models.

ارز ديجيتالي , پيش بيني , تحليل احساسات , يادگيري ماشين

Cryptocurrency , Prediction , Sentiment Analysis , Machine Learning

Fatemeh Feizian

Dr. Babak Amiri