... participants were encouraged to investigate alternative approaches in summarization and report their results NeATS participated only in the fully automatic multi-document summarization task A ... 5.2 Quality Table shows the macro-averaged scores for the humans, two baselines, and 12 systems We assign a score of to all, to most, to some, to hardly any, and to none The value assignment is ... most and above, for some and above, and for hardly any and above They then proceeded to compare system performances at different threshold levels They defined recall at threshold t, Recall t, as...