Estimating Missing Temporal Attributes In Genealogical Data

We present a machine learning approach for estimating missing temporal attributes in genealogical data. Genealogy analyses have been commonly focused on understanding generational relations. The importance of temporal analyses has often been suppresed in genealogical research. We have observed that temporal attributes of an individual, birth, death, marriage and divorce dates, are frequently missing in genealogical data. Filling out those attributes allows users to reconstruct temporal streams of their family stories, which in turn make fruitful genealogy analyses possible.