With the massive surges of new malware, the ‎intuitive detection techniques currently used in most ‎security tools deem ineffective. Consequently, we ‎urgently need better solutions that are established on ‎solid theoretical basis. It becomes, thus, necessary to ‎search for more efficient techniques and algorithms as ‎well as taxonomies and models for attacks and ‎malware. We present, in this paper, the analysis we ‎made on both automatic malware and human-centric ‎attacks, which allowed us to construct a model for ‎attack process. The main objective of this work is to ‎construct a model that can aide in the generation of ‎real attack scenarios and use it in the evaluation of ‎Intrusion Detection Systems. However, the model ‎described here could have many other potential uses. ‎For example, it can be used for writing “execution-‎based” signatures, event correlation, penetration ‎testing, security simulations as well as security ‎educations.‎