Anonymizing Driver and Passenger Data

Privacy is a vital subject for organizations since sharing data with partners and releasing data to the public have become commonplace activities. Secondly, data privacy regulations put forth by governments are to be complied by private organizations.

Under such regulations, companies are expected to draft a data retention policy which should include details regarding data retention time limit, and reasons for retention, to name a few. In a scenario where the data retention time period elapses, organizations are expected to perform one or more of the activities such as

1. returning the data to the concerned individual;
2. transferring the data to another person on the instructions of the concerned individual;
3. destroying the document or
4. anonymising the personal data. Among these four activities, data anonymization is a lucrative option for organizations since anonymized data could be used for historical data analysis and predictive modeling.

In this research collaboration, NTU and GrabTaxi Holdings Pte. Ltd. aims to implement a data anonymizing module with specific focus on the Driver and Passenger data classes. The objectives of this project are

1. conceptualize state-of-the-art data anonymization algorithms suitable for the two aforementioned Grab data classes,
2. implement the conceptualized algorithms as a Python package and
3. evaluate the algorithms with real datasets provided by Grab.