The emergence of anonymity services and the dark web presents a complex landscape where user privacy must be balanced with the need to combat illicit activities. To address this challenge, a Python-based dark web-watching crawler was developed using the Tor network to gather and store a large volume of dark website addresses. This system leverages the Scrapy framework for web crawling to extract specific crime-related information from dark websites, which is then stored in a MongoDB database for analysis. The crawler algorithms analyze the collected data to identify patterns of criminal behavior, and the results are visualized using word cloud diagrams and histograms to create an intuitive interface for real-time monitoring of dark web crimes. This monitoring system aims to enhance law enforcement's ability to detect and respond to illegal activities on the dark web, contributing to efforts to combat human trafficking, illegal information transactions, and the sale of drugs and firearms. The development of this system underscores the importance of leveraging technology to address the dual challenges of user privacy and criminal activity in anonymous online environments. Several machine learning models such as Naive Bayes, decision tree, logistic regression, random forest, KNN were tried for classifying the data from legitimate and dark net websites. Highest accuracy 99% was obtained with Random Forest algorithm.