Research on Interactive Perceptual Wearable Equipment Based on Conditional Random Field Mining Algorithm

Text classification in the natural language processing, information organization, content filtering and other fields have a wide range of applications. The traditional K nearest neighbor method has the advantages of simplicity, robustness, no parameters and high classification accuracy, but it needs to calculate the distance between a new text and all training texts. Therefore, it requires a lot of computation time. Now the mainstream e-commerce site or simply based on the user's rating to determine a comment is a good comment or negative feedback, and this method will often judge the error. This paper presents a method of book review classification based on emotional words. The basic idea is used to divide it into positive or negative according to the number of positive and negative words, which contained in a comment. This paper presents a bootstrap method for emotion word mining based on conditional random field, and divides the newly discovered emotion words into positive and negative categories. Finally, based on emotional words it will be divided into praise and commentary comments.