 This paper presents a novel approach for identifying the most relevant tweet corpus for any given search query. It uses Word2Vec, a neural network-based embedding technique, to measure the relative meaning of words over time by analyzing the changing semantics of words in relation to a particular event. Such as a hurricane, the paper demonstrates how to score tweets based on their dynamic contextual relevance. Additionally, the paper shows how to tune various parameters, including word window size, minimum word frequency, hidden layer dimensionality, and negative sampling, to optimize model performance. This article was authored by Frederick Brown Biggers, Somiya D. Mohanti and Prashanti Manda.