In the field of nuclear and radiation engineering, a huge amount of experimental data related to the reactor physics and the radiation shielding have been obtained at various facilities in the world, and some of them have been released to the public as open data to validate numerical tools solving the reactor physics and radiation shielding problems. Experimental data which can be utilized to validate nuclear data are referred to as integral data in the field of nuclear data, and so many integral data have been accumulated through projects such as ICSBEP and IRPhEP. Those integral data have been efficiently utilized to validate the evaluated nuclear data files.
The number of the available integral data has become huge, so it is important to choose a proper set of integral data when validating the evaluated nuclear data files. To do so, dependency of integral data should be carefully examined, and a procedure how to choose proper integral data should be invented. We have proposed to adopt the concept of the active sub-space to this problem in our previous study and have tested our method against a set of fictitious integral data in the past. In the present work, results newly obtained with actual integral data will be presented.