This task, containing difficulty conditions for the Nogo stimuli, enabled the acquisition of aware errors (mainly in the easy and intermediate conditions) and unaware errors (in the difficult condition), in addition to hits (i.e. correct responses on Go trials).
In summary, this binary classification task contains 158 'experts-annotated' positive example and 2119 partial positive (with 552 from 'groupI' and 1567 from 'groupII') PPI pairs (after removing those pairs labeled as 'not interact' and 'interact' from the experts).
Each condition of the task contained 30 sentences.
The original signal of each task contains 64 non-zero elements at random locations.
Each map task contains a circular memory buffer (with default size 100 MB) where the output data are kept.
Although there are many keys in a map tasks' output, the piece sent to the reduce task contains only one key and its values.
Thus, our reference collection to evaluate tweet contextualization task contains the following: A collection of relevant conversations: these are conversations that are used as a resource to extract components (tweets) of a context corresponding to the contextualization of a tweet.
A single job task contains only one job, while a multiple jobs task contains numbers of jobs.
