Mood Prediction On Tweets Using Classification Algorithm

Data mining is a technique which offers the computer algorithm to compute patterns and find the category of data using classification and clustering. In data mining classification is performed with supervised learning and unsupervised learning. Selection of algorithm depends upon the type and behavior of data. The data can be as structured and unstructured. Structured data is that which reside in fixed field. It is first depends on creating data model. Unstructured data refers to information that does not have a predefined data model or not organized in a predefined manner. In data mining text mining has become an important research area. Text mining is a discovery of new, previously unknown information by automatically extracting information from different resources [5]. The various application in text mining are information retrieval, machine learning, data mining, and statics and computation semantics. In form of text data most of the information is stored. Now a days in a direction of multiple language support most of the research is progressing. This system is capable to group the similar data from different kinds of language source according to their original semantic and also being able to gain information across language [2]. In the presented work the identified twitter data set is used to perform text analysis. Therefore the entire input data samples are required to classify in two classes namely positive and negative. Therefore a binary classifier namely ID3 decision tree and their improved variant is utilized for analysis and performing the classification task. Before classification of text data there is need to improve the quality of data. Therefore the raw text data is first pre-processed then tagged according to the lexical means. After tagging on the original text data the classification algorithms are trained and make use to classify the text according to their sentiments. The implementation of the improved ID3 text classification technique and their performance is evaluated in terms of their accuracy and the error rate. These parameters show how accurately the text patterns are identified using the data mining technique. Additionally for finding their performance in terms of their efficiency the time and space complexity is also measured that shows the effective classification with less consumption.