In order to develop a close relationship between humans and robots, we proposed a multi-modal sentimental model which considers both long and short term affective parameters of interaction. Our model is inspired from scientific studies of love in humans and aims to generate a bi-directional love between humans and robots. We refer to this sentimental connection as "Lovotics" . We have formulated probabilistic mathematical models for identified factors of love, and aim to provide a clear, distinct and discrete interpretation of the intimacy between humans and robots. Such mathematical models are assembled by a Bayesian Network depicting the relationship between intimacy and the causal factors for love. Furthermore, a novel affective state transition system is proposed which takes into account not only the current state caused by interactions, but also the effects of the previous states and internal factors of the robot. Hence, the robot is capable of acting consistently and naturally. The behavior of the robot is controlled by the above two modules via an Artificial Neural Network to develop a realistic affective communication between a human and a robot.