china_eff_clustering.pdf

One-Pass-KMeans-Algorithms/result at master · ryokugyu/One-Pass-KMeans-Algorithms

OKA :

Sort based on qids (not clear)

Picking random seeds and creating the clusters, each time also updating the centroid

Finally adjust the records that are not satisfying k anonymity

Drawback if generalised based on centroid only is → too much loss of data followed in this paper