High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Weather Analysis using Twitter Data


Elton Asher Jose Menezes , Agnel Institute Of Technology And Design; Aniket Arvind Patil, Agnel Institute Of Technology And Design; Akash Chaware, Agnel Institute Of Technology And Design; Abid Ghori, Agnel Institute Of Technology And Design; Shreedatta Sawant, Agnel Institute Of Technology And Design


Naive Bays Classifier, K-Nearest Neighbor


Weather is an important aspect of our human lives. Human can adapt to different climatic conditions but they also need to dress appropriately for them; so the weather conditions should be known to them. Weather Analysis using Twitter data aims to gather a series of Tweets based on a particular location and generate the weather condition/ conditions that the Tweets are making a reference to. An Unsupervised Learning Approach will be used in-order to automatically classify the Tweet/Tweets of a particular location into appropriate weather categories i.e. Sunny, Windy, Cold, Humid, Hot, Chilly, Cool. To achieve Unsupervised Learning Approach a Data Set will be given to a Classifier , like Naive Bayes/K-Nearest Neighbor in-order to determine which weather conditions the tweet or set of tweets belong to and also to help the classifier improve its accuracy in determining the correct output overtime. In order for classification, pre-processing steps will be carried out on the Tweet/Tweets obtained. The pre-processing steps consist of Removal of Stop Words, Removal of Hashtags, Removal of '@'symbol etc. The Steps will be explained in detailed throughout this document. Training of the Classifier involves having a set of Tweets that are already classified into the appropriate weather conditions. Once the Classifier has been trained; the test data is given to the classifier to determine the appropriate categories that the Tweet/Tweets fall under and what is the weather condition that the person/group of people belonging to that particular location are talking about.

Other Details

Paper ID: IJSRDV6I40818
Published in: Volume : 6, Issue : 4
Publication Date: 01/07/2018
Page(s): 1023-1027

Article Preview

Download Article