Tamr, Inc.

George Beskales

Ihab Francis Ilyas

Daniel Meir Bruckner

Nikolaus Bates-Haus

Michael Ralph Stonebraker

Alexander Richter Pagan

Liam Cleary

Paul Roome

John Kraemer

Jason Liu

Stan Zdonik

Andrew H Palmer

Aizana Turmukhametova

Vladimir Gluzman Peregrine

Tianyu Zhu

Kanak Kshetri

Given a number of records and a number of target classes to which these records belong to, a (weakly) supervised machine learning classification method leverages known possibly dirty classification rules, efficiently and accurately learns a classification model from training data, and applies the learned model to the data records to predict their classes.

Computer-implemented method for performing hierarchical classification

An end-to-end data curation system and the various methods used in linking, matching, and cleaning large-scale data sources. The goal of this system is to provide scalable and efficient record deduplication. The system uses a crowd of experts to train the system. The system operator can optionally provide a set of hints to reduce the number of questions send to the experts. The system solves the problem of schema mapping and record deduplication a holistic way by unifying these problems into a unified linkage problem.

Method and system for large scale data curation

Fast record deduplication is accomplished by providing as an input, data records having multiple attributes, and local similarity functions of individual attributes with local similarity thresholds. Bin IDs are then generated based on the local similarity functions and the local similarity thresholds. The Bin IDs are unique identifiers of a respective bin of records, and the bin of records is a set of records that are possibly pairwise similar. Local candidate pairs are identified based on data records that share Bin IDs. The local candidate pairs are aggregated to produce a set of global candidate pairs. The set of global candidate pairs are filtered by deciding whether a pair of data records represents a duplicate.

Scalable binning for big data deduplication

A data curation system is provided that includes various methods to enable efficient reuse of human and machine effort. To reuse effort, various facilities are presented that model, save, and allow for querying of provenance and state information of a curation workflow and allow for incremental, stateful transitions of the data and metadata thereof.

Data curation system with version control for workflow states and provenance

An end-to-end data curation system and the various methods used in linking, matching, and cleaning large-scale data sources. The goal of this system is to provide scalable and efficient record deduplication. The system uses a crowd of experts to train the system. The system operator can optionally provide a set of hints to reduce the number of questions sent to the experts. The system solves the problem of schema mapping and record deduplication in a holistic way by unifying these problems into a unified linkage problem.

A very efficient computer system is presented to generate all pairs of records that have a certain similarity. Similarity is defined in terms of the textual similarity of the record attributes and/or absolute difference for numeric record attributes. Software assigns each record to a number of bins, and then compares pairs of records that belong to the same bin. This is more efficient than comparing all pairs of records since the number of records compared to each other is much smaller.

Growing community of inventors

Waltham, MA, United States of America

George Beskales