Despite the commercial success of Location-Based Services (LBS), the sensitivity of the data they manage, specially those concerning the user's location, makes them a suitable target for geo-location inference attacks. These attacks are a new variant of traditional inference attacks aiming at disclosing personal aspects of users' life from their geo-location datasets. Since this threat might dramatically compromise the privacy of users, and so the confidence of LBS, a deeper knowledge of geo-location inference attacks becomes essential to protect LBS. To contribute to this goal, this short paper makes a step forward to model well-known types of geo-location inference attacks as a previous step to quantitatively assess the privacy risk they pose. © 2014 IEEE.