Abstract: Systems and methods classify pollutants based on multifactor analysis of data from sensors in a monitored area and contextual data from remote or local sources. Classifying a pollutant in air at a monitored area may include operating a particulate matter sensor to produce raw data representing measurements of particulate matter in the air, evaluating pulses in the raw data to determine a pulse width and a maximum for each pulse, and identifying a type for the pollutant in the air using a classification model and data including the pulse widths and the maxima of the pulses. The data use in classification may further include non-particulate measurements from local chemical or environmental sensors and contextual data from the cloud or from local user devices.