Websites for microblogging are crucial sources of current information during any disasters, whether they are man-made or natural. Designing and testing Information Retrieval (IR) systems or algorithms that pull information from microblogs is so crucial. It is also necessary to retrieve the information in accurate and fast manner. In this research paper one word embedding based algorithm is proposed. The algorithm is experimented with more than 50000 tweets related to natural disaster of Nepal earth quake. The algorithm is compared with well known machine learning based algorithms such as Decision Tree, Random Forest and Naïve Bayes. The result of proposed algorithm and all other well-known algorithms is measured in precision, recall, F1 Score and execution time. The proposed algorithm excels in all parameters.