This project is testing if a correlation between professional sentiment and the stock price exists, and if it does, how do they correlate. The project will try to generate sentimental indicators from financial news as independent variables, and use the corresponding market index as the training dataset.
I think this project is very interesting since it is looking at a practical issue that will truly benefit the financial traders. Because investigation about the tweets’ sentimental implication has showed correlation, it makes sense to want to know more about if professional reviews, instead of public reviews can be a better variable in predicting stock trends. The project proposal is very clear about what they want to do.
However, I do have some concerns and suggestions towards the project proposal. First of all, it will be nice if the proposal or the final report emphasize more on what benefits can the success of this project brings. Secondly, I am a little worried about how the sentimental indicators are generated. Since the proposal only mentioned using financial news, natural language processing can require really large training sets and may sometimes return inaccurate results if the texts are complicated. Also, I think the team should mention if there are other factors that can influence their results, and their methods for the generation of their indicator variable. Since many of the news articles are correlated and share the same content, some of the predictors can be dependent on each other, and this may influence the results of this test.
This project is testing if a correlation between professional sentiment and the stock price exists, and if it does, how do they correlate. The project will try to generate sentimental indicators from financial news as independent variables, and use the corresponding market index as the training dataset.
I think this project is very interesting since it is looking at a practical issue that will truly benefit the financial traders. Because investigation about the tweets’ sentimental implication has showed correlation, it makes sense to want to know more about if professional reviews, instead of public reviews can be a better variable in predicting stock trends. The project proposal is very clear about what they want to do.
However, I do have some concerns and suggestions towards the project proposal. First of all, it will be nice if the proposal or the final report emphasize more on what benefits can the success of this project brings. Secondly, I am a little worried about how the sentimental indicators are generated. Since the proposal only mentioned using financial news, natural language processing can require really large training sets and may sometimes return inaccurate results if the texts are complicated. Also, I think the team should mention if there are other factors that can influence their results, and their methods for the generation of their indicator variable. Since many of the news articles are correlated and share the same content, some of the predictors can be dependent on each other, and this may influence the results of this test.