2012 ©
Journal Publication
Title of Article Topic Modeling on Crowd Trading Ideas for Digital Asset Price Prediction 
Date of Acceptance 5 January 2023 
     Title of Journal International Journal of Applied Engineering & Technology 
     Standard SCOPUS 
     Institute of Journal Roman Science Publications  
     ISBN/ISSN ISSN: 2633-4828 
     Month January
     Year of Publication 2023 
     Page 6-12 
     Abstract Tradingview is the most widely recognized social network platform for stock and digital asset trading, which a number of investors access to share their thoughts on investments each day. The present study sought to analyze contents published in the “Ideas” forum on Tradingview from November 2021 – October 2022 using topic modeling through Latent Dirichlet Allocation (LDA). The results demonstrated that 9,553 texts drawn from Tradingview’s Ideas were classified into six topics about digital asset trading. Based on the hypothesis and the prediction, it was found that the most shared idea, particularly 3,354 texts, concerning trading strategy and risk/profit, followed by 3,101 about trend line, high/low price pattern, and support/resistance. The number of texts on both topics accounted for 67.56%. It can be implied that investments in digital assets require good strategies, risk management, analysis of price trends, and knowledge of support and resistance. Identifying significant words and topics using LDA uncovers a larger number of words latent in the texts than hypothesiz. 
     Keyword digital assets, crowd idea user, topic modeling; Latent Dirichlet Allocation, digital trading, trading ideas  
607080012-9 Mr. SEKSAK PRABPALA [Main Author]
Humanities and Social Sciences Doctoral Degree

Reviewing Status มีผู้ประเมินอิสระ 
Status ตีพิมพ์แล้ว 
Level of Publication นานาชาติ 
citation false 
Part of thesis true 
Attach file
Citation 0