UW-Madison CS839 Course project under Theo Rekatsinas.
This project deals with the design and applications of algorithms to perform embedding of records from a database. We aim to draw parallels with the methods currently used for word and sentence embedding and explore applications of such an embedding scheme, namely detection of erroneous records and error correction. Here we present an approach to embed records and then study its applications by building an end to end pipeline to perform error detection and error correction of records.
- Error detection
- Imputation of missing data
- Error correction