This is an estimation of the number of people in your community who are infectious as a fraction of the total population of your community. This parameter is not easy to quantify accurately, but one can hope to get the order-of-magnitude right from the disease prevalence data and/or the epidemiological models. Remember that infectious/contagious is different from infected numbers. People are thought to be contagious mostly the week around the onset of symptoms, so that has to be taken into account in the estimates. Also, there is a fraction of undetected contagious cases (asymptomatic / presymptomatic), which will increase transmission. Plus one would hope that a major fraction of the cases that are in quarantine or a hospital and not transmitting the disease much. The uncertainty on the fraction of contagious individuals in the community is one more reason why the absolute risk values will be uncertain, but the relative risks will still be robust. An in-between value of 0.3% is used as default, as the approx. geometric mean of the results below.
Data would be specific to your own region. For example, for the US, for different counties, you can get some quick estimates from this webpae.