When I was taught in my undergrad, we used the rule of thumb that if the difference between the temperature and dewpoint was less than about 2 degrees celsuis in the observed sounding, then you have cloud.
The National Weather Service forecasters I know in the US use infrared satellite imagery and areas of relative humidity greater the 80% in the forecasts to 'suggest' clouds.
Geeks like me who spend all day messing with the numerical models can get these things to spit out concentrations of hydrometeors in the numerical forecasts/analyses, which should be a bit less vague (assuming the models are correct). In the US, the government agencies supply this data to the public e.g. this page:
http://www.ral.ucar.edu/weather/model/ but in the UK and Europe the wise government think they can make wads of cash selling this info back to the people who have already paid for it.
Cheers,
Gareth.