Webseite Deutsches Zentrum für Luft- und Raumfahrt (DLR) - German Aerospace Center
Millions of Tweets generated every second by human contributors contain insightful information regarding various events constantly happening around us. The high temporal resolution of these data make them invaluable for near real-time observation of events.
Within this project the student will get hands-on experience with handling and utilising the Twitter Live stream as well as static Twitter datasets to harvest and localise most suitable Tweets for a given disaster event. Further, the student will conduct various analytical tasks by using machine learning algorithms together with Natural Language Processing (NLP) algorithms to explore its linguistic features to help detect disaster events more accurately, e.g., by using Named Entity Recognition (NER) or Hidden Markov Models (HMM). The student will further utilise these results to verify events observed through EO data. This topic can be developed with a study project, and can be further continued on to a thesis.
Um sich für diesen Job zu bewerben, sende deine Unterlagen per E-Mail an Dr.HansiSenaratneemail@example.com