| Date: | 2015, June 24 |
| Time: | 11:00 |
| Author: | Engel, Christoph |
| Title: | Information Extraction for Traffic Management |
The traffic management needs a lot of structured data from different sources e.g. for navigation depending on the current traffic situation or to calculate prognoses of the upcoming traffic situation in a town. All these Applications are dependent on a good quality of the incoming (sensor-) data to achieve good results. But not all data sources will provide high quality data. Furthermore the configuration of some Applications is based on certain assumptions which may not always apply. In these cases the Applications will generate incorrect results. On the other hand a large amount of traffic information is distributed on communication platforms like Twitter and Facebook. So the idea is to convert these unstructured data into structured data and use it e.g. to extend classical traffic information, verify the results generated by an Application or even use it as an additional data source.
The presentation covers an overview of the state of the art in natural language processing compared with the requirements in traffic management as well as a first idea how to make messages from Twitter usable in traffic management.