Machine Learning and Detection of Impersonated Domains

We have html source codes and screenshots for domains. We need to build a classifier which categories the domains as parked/other domains and then later use the data to build a model which correctly predicts whether the domains are malicious or not (I have some experience in ML so I will be assisting). Also, please know that there will be adhoc requests like automating some processes which have to be done. Please understand that this project is not for a day or a week. There will be simple tasks spread over 1 or 2 months. Kindly bid only if you can consider this.