Resolving conflicting observations and non-binary claims
Abstract
As we discussed in previous chapters, there are some limitations of the basic maximum likelihood estimation (MLE) model we introduced in Chapter 5. Two of them are: (1) the reported observations from different sources on the same claim are assumed to be corroborating. In another word, they do not contradict with each other. (2) The claims are assumed to be binary only. However, the above assumptions may not always hold in various kinds of social sensing applications. Interesting questions emerge: how to generalize the MLE model to explicitly handle conflicting observations from sources and non-binary claims? In this chapter, we reviewed two generalizations of the ...
Get Social Sensing now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
