Diagnosing Alternative Facts

This paper presents an approach to applying model-based diagnosis to the task of interpreting information from a wide variety of sources: text, video, meta-data, audio, etc. Much of the information streams are contradictory, incomplete, purposely deceptive or biased. People have to make critical decisions based on such murky information. The ultimate users of such systems range from intelligent analysts to ordinary folk. We have developed a preliminary version of our HCDX tool (hypothesis construction through diagnosis). We plan to distribute this tool as open source.

Citation

de Kleer, J.; Klenk, M.; Feldman, A. Diagnosing Alternative Facts. DX 2017 : 28th International Workshop on Principles of Diagnosis.; Brescia, Italy. Date of Talk: 9/26/2017