An approach to searching for anomalies in concrete historical data

The article continues the series of works devoted to the technology of concrete historical research supporting. The technology is based on the principles of co-creation and crowdsourcing and is designed for a wide range of users which are not professional historians and biographers. The article is devoted to the further development of the technology by integrating the mechanism of automated search for anomalies in concrete-historical information. The analysis of the existing approach to the search for contradictions in historical and biographical facts has been carried out and its limitations and shortcomings have been revealed. The expediency of the transition from the search for obvious contradictions to the search for anomalies is justified within the development of this approach. The causes of anomalies have been analyzed, their classification has been carried out, and the features of anomalies in concrete-historical data have been determined. The analysis of the known methods for point anomalies detecting has been carried out and the impossibility of using the methods based on supervised learning as well as metric methods in the technology has been justified. The most promising method based on clustering has been found and further steps for its implementation have been determined. The necessary changes to the technology object model are described and justified.