In this paper, we present an operational system for cyber threat intelligence\ngathering from various social platforms on the Internet particularly sites on\nthe darknet and deepnet. We focus our attention to collecting information from\nhacker forum discussions and marketplaces offering products and services\nfocusing on malicious hacking. We have developed an operational system for\nobtaining information from these sites for the purposes of identifying emerging\ncyber threats. Currently, this system collects on average 305 high-quality\ncyber threat warnings each week. These threat warnings include information on\nnewly developed malware and exploits that have not yet been deployed in a\ncyber-attack. This provides a significant service to cyber-defenders. The\nsystem is significantly augmented through the use of various data mining and\nmachine learning techniques. With the use of machine learning models, we are\nable to recall 92% of products in marketplaces and 80% of discussions on forums\nrelating to malicious hacking with high precision. We perform preliminary\nanalysis on the data collected, demonstrating its application to aid a security\nexpert for better threat analysis.\n