Mahmood Tariq

Advisor

Sergio Greco

Co-advisor

Gianvincenzo Alfano

Research Topic

Data Imputation for incomplete data

Research Abstract

Data integration is an essential task in today’s world, as it is necessary for organizations, institutions, and businesses to combine data from multiple sources. Record linkage (a.k.a. Entity Resolution) is the process used to identify records, possibly coming from multiple databases, referring to the same entity. When applied on a single database, the process is known as data deduplication. Record linkage for incomplete or incompletely-matched records has been studied in recent research. The current research focuses on data pre-processing to improve data quality. To this end, different machine learning and deep learning algorithms have been studied to achieve the desired objective