Women safety is a crucial issue across the globe and social media has grown to be significant platform as it provides space for the individuals to share their thoughts, insights, concerns and express their opinions, experiences, and concerns on this issue. The examination of tweets pertaining to women’s safety might yield significant understanding of the attitudes, patterns, and challenges women encounter in society. The paper focuses on implementing machine learning frame work to analyze tweets related to women's safety and classifying the sentiments in to positive, negative, and neutral categories. The analysis took into account the twitter data of the major Indian cities collected with decision tree, k-nearest neighbor, logistic regression, random forest algorithms. The results evaluated with the performance metrics accuracy, F1 score, precision, and recall indicated the better performance of kNN model that provided an accuracy of 89% with F1 score, precision, and recall of 0.92, 94.4%, and 89.9%, respectively. The analysis can aid activists, organizations and legislators create focused plans, interventions and public awareness to address the difficulties women encounter in different situations.