Advances in haptic technology have led researchers and engineers to seek out killer applications in which users can enjoy an experience of touch in AR/VR spaces. Such applications will respond appropriately to human desire for haptic experiences (i.e., touch desire) and thus it is essential for researchers and engineers to understand the nature of people's touch desires as they arise in the course of daily life. In this study, we employed Twitter data analysis to investigate a diurnal pattern in touch desire. Our results showed that touch desire identified in and extracted from Twitter texts did reveal a diurnal pattern. Touch desire tended to be at its lowest in the morning and increased as the day progressed. The time at which it peaked varied with the specific target of touch desire. Touch desire in relation to other people and objects reached its peak at night, but touch desire in relation to animals reached its peak at noon. These results were confirmed not only by our Twitter text analysis but also by data from other social media and an online survey. In addition, we found that the diurnal pattern of touch desire for each target shows a strong correlation with that of visual desire for the same target. This suggests that the diurnal pattern of touch desire is not limited to the sense of touch but is common to other sensory desires for each target. Our findings suggest that researchers need to take the time of day into account when investigating touch desire. Our findings also offer valuable insights for developers into the design of haptic applications and displays that takes into account the timing of daily peaks in touch desire.