Machine Learning, Sentiment, and Social Media Analysis for Financial Analytics

点击翻译按钮获取中文摘要

Social media platforms such as Twitter or StockTwits are widely used for sharing advice between investors,**traders, and entrepreneurs. The popularity of these platforms, the vast amount of information generated by**active members, and their usage by influential people imply that this social media content is often predictive of**financial markets. To leverage this observation, Periscope Capital has developed methods for monitoring financial**market discussion on social networks to inform its portfolio investment and trading decisions.**In this project, Periscope Capital will partner with the University of Toronto to develop novel algorithms to model**the influence of social media on stock market performance. These novel methods are based on machine learning**principles that will provide financial analysts with powerful tools for summarizing social media sentiment and**opinion regarding individual stocks as well as providing quantitative predictions for use in investment and trading**decisions. The proposed tools will also help identify influential social media users as well as influential social**media posts. The proposed innovations are crucial for Periscope Capital as they compete with other hedge fund**companies and will enable the development of innovative and novel solutions to maintain a high-performance**investment portfolio that will benefit their business and investors across Ontario and beyond.