In this work the derived occurrence probability of solar energetic particle (SEP) events (i.e. proton events measured at Earth’s position) and their peak fluxes and total fluences depending on coronal mass ejection (CME) parameters, i.e. linear speed (V) and the angular width (AW) are presented. A new SEP catalogue with associated CME data from 1997 to 2013 is utilized. It is found that the SEP probability strongly depends on the CME speed and the angular width as follows: The highest association (72.70%) is obtained for the full halo CMEs with V ≥ 1500 km s−1 and the lowest association (0.7%) is found for the non halo CMEs with 400 km s−1 ≤ V ≤ 1000 km s−1. The SEP occurrence probabilities are different as much as 26 times according to the CME speed (V), comparing fast versus slow CMEs and 44 times according to the AW, comparing halo to non halo CMEs. Furthermore, linear regressions of the proton peak flux and integral fluence at several integral energy channels (E > 10 MeV, E > 30 MeV, E > 60 MeV, E > 100 MeV) were obtained. Our results, were used to build a module of an operational forecasting tool (i.e. FORecasting Solar Particle Events and Flares – FORSPEF, http://tromos.space.noa.gr/forspef/). This module performs nowcasting (short term forecasting) of SEP events using near real-time CME identifications obtained from CACTus (http://sidc.oma.be/cactus/). The outputs offered by the operational module of the tool to the end user (textural, pictorial, archived data) are presented. Finally, the validation of the system, in terms of archived data is described, in terms of categorical scores (Probability of Detection – POD and a False Alarm Rate – FAR).