The Jiangmen Underground Neutrino Observatory (JUNO) is the largest underground liquid
scintillator experiment in the world, currently under construction in southern China. JUNO aims
to determine the neutrino mass ordering as its primary experimental goal by measuring the energy
spectrum of reactor neutrinos. Its central detector consists of 20-kton liquid scintillator and
more than 17,000 20-inch photomultiplier tubes. Vertex reconstruction and particle identification
capability are important components in the event selection to suppress the backgrounds of the
reactor neutrino events. The former is also crucial to understand the non-uniform energy response of the detector to achieve 3%/ 𝐸 [MeV] of the energy resolution. In this proceedings paper, a data-driven method of reconstructing the event vertex position as well as separating positron and 𝛼/fast-neutron is discussed. Both of the algorithms can be developed with the radioactive neutron (Americium-Carbon) source which is planned to be regularly deployed to calibrate the JUNO detector responses. We present the detailed methodology and reconstruction performances using the JUNO detector simulation.