Place Detection

by KAIST

Detecting places in which the users have visited

Our project aims to develop a highly accurate algorithm for detecting significant places in which the users have visited. We sure that this place detection algorithm will be beneficial for other location-based services, which have used the raw coordinates only.
Based on the data that we have collected, a paper for our algorithm is recently published in IEEE Transactions on Mobile Computing, Vol 12, Issue 5. Furthermore, we want to validate our algorithm in the world-wide data sets provided by OpenPaths.
The place detection service for smartphones is also being developed. We hope that this useful service can cooperate with OpenPaths.

Sequences of coordinates collected from each anonymous user is inputted to our algorithm.
The algorithm will generate a list of possible places in which the user have visited.

We will securely save the data in an well-protected PC.

The title of the paper published in IEEE Transactions on Mobile Computing is "A Probabilistic Place Extraction Algorithm based on a Superstate Model", and more papers will be submitted with the successful evaluation results with OpenPaths.
Also, place detection service will be implemented, and we hope that the service can cooperate with OpenPaths.