Numerous attempts have been made to estimate position using Wi-Fi signals. Radio map, which is the collection of the received signal strength indicator (RSSI) of Wi-Fi signals along with their collected location information, is essential for the estimation of positon in a high resolution. However, the radio map construction is not only labor intensive, but also it requires periodic updates to cope with the changes of Wi-Fi environments. This paper proposes a practical location-labeling method for crowdsourced fingerprints to construct Wi-Fi radio maps in a large shopping mall environment. In the first step, the initial radio map is constructed based on a small number of reference data obtained from mobile payment transactions. In the second step, the path of the crowdsourced fingerprint sequence is accumulated and then their collected locations are estimated to improve the radio map. Experiments performed at a landmark building revealed the proposed method was effective in location-labeling of crowdsourced fingerprints.