I'm trying to develop an SVM model to classify some texts into positive and negative. For the training phase, I'm going to use a dataset containing 2400 comments from Twitter, classified into positive and negative. I would ask you how can I do a scatterplot of this dataset in order to see the data distribution and understand whether it's linear or non linear, so I can choose whether apply the kernel trick or not (or whetehr to apply another algorithm instead of SVM). The dataset is made of only of the attributes "text" and "sentiment", so do I need to calculate other attributes such as polarity, frequency or anything else to create the scatter plot? An which attributes should I use as coordinates of the plane? For my project, I'm using Rapidminer and Python, but I appreciate any other suggestion to visualize the data.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.