The speed of change in the modern world is impressive. Within the last 50 years, many devices and technologies have significantly transformed their appearance, intrinsic characteristics and improved their performance. Computers have changed from the size of several cubic meters to thin laptops. Photo and video camera, recorder, computer, and phone are not anymore five devices but have converged to one small smart device that anyone can put in the pocket. Autonomous vehicles can drive and park themselves. The progress did not disregard such industrial fields as medicine, robotics, energy harvesting, and others looking for new materials and technologies with breakthrough potential. Dielectric Elastomers (DEs) are versatile and smart materials that operate in three modes as actuators, generators, or sensors. Actuators change their form in response to an applied electric field. They could be used as artificial muscles in biomedical applications. As generators, DEs can convert mechanical energy into electricity. Finally, they can also be capacitive sensors. Among all DEs, polysiloxanes possess all requirements for constructing robust and reliable devices, including artificial muscles that can outperform the characteristics of natural muscles in terms of stress and strain. However, the performance of polysiloxane-based devices is restricted by their low dielectric permittivity value ('~2.8). A higher dielectric permittivity would increase sensor sensitivity, the energy generated, or reduce the actuator operation voltage, which is typically in the kV range. Chemical modification of polymers with polar groups allows for an increase indielectric permittivity but may also increase the glass transition temperature, which would ideally remain significantly below room temperature. However, a systematic investigation on how different polar groups and degree of functionalization alters the dielectric and thermal properties is missing. This thesis presents a systematic approach to polysiloxanes modified with different types and contents of polar groups and explores the potential of such polar silicones as dielectric materials in transducers. To this end, we set out from polysiloxanes with different vinyl group content. These groups were then transformed into various polar groups by an efficient one-step thiol-ene addition post-polymerization modification. The resulting collection of novel materials was ideally suited for the first step towards elucidating structure-property relationships, which are essential for finding the optimum material. It has allowed us to select the most promising polar polysiloxanes for a defined application. Next, a strategy was developed to cross-link the most promising polar polysiloxane to thin-film elastomers with useful mechanical and electromechanical properties. An incomplete functionalization of the vinyl groups on polymethylvinylsiloxanes with less than stoichiometric amount of thiols allowed us to obtain polymers with around 2.5% unreacted vinyl side groups. They were subsequently cross-linked into thin films with tuneable properties via the fast and reliable thiol-ene addition. Finally, single membrane DE actuators were constructed, and their electromechanical performance was tested. The best materials showed around 7.5% actuation strain at only 300 V, while a visible actuation at 200 V was also detected. Actuators that can be operated for more than 50'000 cycles were obtained with some DEs.