The CrisisLex.org initiative offers a variety of datasets and tools relating to the use of social media for crisis and disaster response. Two of their topical taxonomies, "CrisisLexRec – Recommended CrisisLex Lexicon" and "EMTerms – Terminological Resource for Crisis Tweets" are now available within the GDELT GKG (prefaced with "CRISISLEX_". Both of these lexicons were built specifically for use with social media (Twitter in particular) and thus do not map perfectly onto the language used in traditional mainstream media.
You will likely wish to combine mentions of "CRISISLEX_CRISISLEXREC" with the other categories in order to narrow their scope:
- CRISISLEX_C01_CHILDREN_AND_EDUCATION
- CRISISLEX_C02_NEEDSPROVIDE_FOOD
- CRISISLEX_C03_WELLBEING_HEALTH
- CRISISLEX_C04_LOGISTICS_TRANSPORT
- CRISISLEX_C05_NEED_OF_SHELTERS
- CRISISLEX_C06_WATER_SANITATION
- CRISISLEX_C07_SAFETY
- CRISISLEX_C08_TELECOM
- CRISISLEX_O01_WEATHER
- CRISISLEX_O02_RESPONSEAGENCIESATCRISIS
- CRISISLEX_T01_CAUTION_ADVICE
- CRISISLEX_T02_INJURED
- CRISISLEX_T03_DEAD
- CRISISLEX_T04_INFRASTRUCTURE
- CRISISLEX_T05_MONEY
- CRISISLEX_T06_SUPPLIES
- CRISISLEX_T07_SERVICESNEEDEDOFFERED
- CRISISLEX_T08_MISSINGFOUNDTRAPPEDPEOPLE
- CRISISLEX_T09_DISPLACEDRELOCATEDEVACUATED
- CRISISLEX_T11_UPDATESSYMPATHY
The proper citation for the "CRISISLEX_CRISISLEXREC" theme is:
- A. Olteanu, C. Castillo, F. Diaz, S. Vieweg. CrisisLex: A Lexicon for Collecting and Filtering Microblogged Communications in Crises. In Proceedings of the AAAI Conference on Weblogs and Social Media (ICWSM'14). AAAI Press, Ann Arbor, MI, USA, 2014
The proper citation for the other themes is:
- I. Temnikova, C. Castillo, and S. Vieweg: EMTerms 1.0: A Terminological Resource for Crisis Tweets. In Proceedings of the International Conference on Information Systems for Crisis Response and Management (ISCRAM'15). Kristiansand, Norway, 2015.
Given the mismatch between social and mainstream media language use, you will likely see a significantly elevated false positive rate for the time being while we observe how the taxonomies map onto mainstream news coverage over the coming weeks in order to fine-tune them for the language of news media reporting. So, stay tuned!