Information from social network sites, blogs and web pages generates massive amount of unstructured data. As this big data is generating at faster rate, analysis part is a challenge to data analysis. To manage and extract useful information from this type of data is a cumbersome job. In this paper, we discuss various machine learning techniques to process and handle this data. The advantages and drawbacks each method is stated based on various metrics such as scalability, real-time processing and data size supported. K-means clustering, SVM, and Genetic algorithms are applied on data to get performance impact of features. A comparison of these machine learning techniques has been performed in order to obtain a reasonable algorithm for detection of useful information from online updating resources such as news and tweets. The best result has been achieved by a k-means clustering.
展开▼
