Abstract:Over the past years, the effect of massive information from the web on the financial market has increased. How to process and utilize such information attracts both researchers and portfolio managers. In this paper, financial information obtained daily from the web is treated as a time series and then associated with stock price volatilities. First, six research topics on financial time series are outlined, namely, analyses of stock price time series P, trading volume time series V, web information time series W, and relationship between P and V, P and W, as well as V and W. Second, a model connecting P and W based on the support vector regression (SVR) is examined as an example of the six research topics. Third, given that a typically successful way of computer-based natural language processing is through the conduct of keyword analysis, the novel finance-computer time series W is explicitly defined in terms of financial keywords and is used in the present paper as the topic of investigation. The relationship between P and W is modeled using SVR. Because during the pre-web era people cannot manually and efficiently process image information from the newspapers and sounds from the television and radio over a longer period of time (e.g., a year), they were unable to obtain the time series W. Therefore, it is the web that makes the research on the relationship between P and W in the meaning of quantity of W possible. Finally, experiments on the Shanghai and Shenzhen security markets revealed that the introduction of W helps improve model accuracies. As the web further develops, more and more ordinary people share their views on the web. The “long-tail” of massive financial information formed by these “grass roots” has a noticeable effect on the financial markets. In financial markets, those who quickly capture and interpret financial information have the potential to generate profits. With the use of the newly found model connecting P with W and fast decision making, financial market practitioners can be rewarded.

Stock Price Forecasting with Support Vector Machines Based on Web Financial Information Sentiment Analysis.

Forecasting Stock Market Movement Direction Using Sentiment Analysis and Support Vector Machine

Stock Market Forecasting Based on Text Mining Technology: A Support Vector Machine Method

Predict financial text sentiment: an empirical examination

Application of Machine Learning With News Sentiment in Stock Trading Strategies

Stock Prices Forecasting and Optimization Strategies Based on Support Vector Machines

A Stock Market Trend Forecasting Method Based on Stock Sentiment Analysis

The role of text-extracted investor sentiment in Chinese stock price prediction with the enhancement of deep learning

Mining Opinion and Sentiment for Stock Return Prediction Based on Web-forum Messages

Financial Forecasting of Stock Market Using Sentiment Analysis and Data Analytics

The Research on Financial Volatility with Sentiment Analysis

Stock Market Prediction Exploiting Microblog Sentiment Analysis.

Sentiment Analysis in Stock Price Prediction: A Comparative Study of Algorithms

Stock Market Prediction Based on Financial News Text Mining and Investor Sentiment Recognition

Stock Prediction Based on News Text Analysis

Incorporating stock prices and news sentiments for stock market prediction: A case of Hong Kong

Stock Market Prediction from Sentiment and Financial Stock Data Using Machine Learning

News impact on stock price return via sentiment analysis.

Stock Price Forecasting by Combining News Mining and Time Series Analysis.

Explainable stock prices prediction from financial news articles using sentiment analysis

Associating Stock Prices with Web Financial Information Time Series Based on Support Vector Regression