Abstract: Understand the time interval that an event is contained is key for different decision making services. For instance, a secure route suggestion needs crime data to identify crime hotspots inside a time window and select safe routes. Time windows help to separate distinct situations and focus the analysis within a time interval. Also, the result may provide an insight into the changes that occur during the day. With this in mind, this paper presents an approach to identify mobile and variable time windows with the goal of discovering hotspots, named MARTINI. The hotspots may be used by different types of services that want the granularity applied in this paper. The data is fragmented with the objective to identify the situation according to each day of the week, data type, and more. MARTINI utilizes a Gaussian Distribution Function to describe the event density of different data types and time intervals. In addition, it uses this representation to find out the changes that occur during the day. The results obtained show that MARTINI requires less time to recognize changes in the situation with a 10 minute sensitivity. In addition, it outperforms the smaller time window even with a 2 hour interval Ver menos