OVERALL ACC: (BEFORE OUTPUT FORMATTER) 16 / 50 = 0.32 -------------------------------------------------------------------------------- OVERALL ACC: (AFTER OUTPUT FORMATTER) 17 / 50 = 0.34 Formatter Errors: [] -------------------------------------------------------------------------------- AGGREGATOR PERFORMANCE: EASY: 16 / 16 = 1.0 HARD: 0 / 19 = 0.0 -------------------------------------------------------------------------------- ANSWER MODULE STATS (WHEN CORRECTLY ACTIVATED ACCORDING TO THE DISPATCHER): ACC Single-Span-Extraction 15 / 25 = 0.6000 Single-Span-Multi-Hop-Extraction 6 / 25 = 0.2400 Multi-Spans-Extraction 0 / 7 = 0.0000 YesNo 2 / 4 = 0.5000 Arithmetic-Operations 0 / 1 = 0.0000 Counting 0 / 3 = 0.0000 Date-Duration 0 / 4 = 0.0000 Kinship 5 / 6 = 0.8333 -------------------------------------------------------------------------------- ANSWER MODULE STATS (WHEN WRONGLY ACTIVATED ACCORDING TO THE DISPATCHER): ACC Single-Span-Extraction 1 / 25 = 0.0400 Single-Span-Multi-Hop-Extraction 1 / 25 = 0.0400 Multi-Spans-Extraction 3 / 43 = 0.0698 YesNo 0 / 46 = 0.0000 Arithmetic-Operations 8 / 49 = 0.1633 Counting 8 / 47 = 0.1702 Date-Duration 18 / 46 = 0.3913 Kinship 7 / 44 = 0.1591 -------------------------------------------------------------------------------- ANSWER MODULE ERROR SAMPLES (WHEN CORRECTLY ACTIVATED): Single-Span-Extraction: ['D033Q04', 'D033Q05', 'D042Q04', 'D096Q05', 'D116Q03', 'D179Q06', 'D241Q04', 'D244Q04', 'D247Q02', 'D312Q05'] Single-Span-Multi-Hop-Extraction: ['D001Q09', 'D033Q04', 'D033Q05', 'D042Q04', 'D042Q05', 'D072Q06', 'D087Q07', 'D090Q01', 'D096Q05', 'D107Q08', 'D116Q03', 'D179Q06', 'D241Q04', 'D245Q01', 'D247Q01', 'D247Q02', 'D247Q05', 'D312Q05', 'D312Q07'] Multi-Spans-Extraction: ['D008Q06', 'D010Q06', 'D037Q05', 'D088Q06', 'D105Q05', 'D215Q07', 'D318Q04'] YesNo: ['D087Q05', 'D106Q05'] Arithmetic-Operations: ['D248Q04'] Counting: ['D015Q06', 'D069Q13', 'D117Q01'] Date-Duration: ['D076Q10', 'D105Q02', 'D310Q05', 'D311Q07'] Kinship: ['D078Q07'] --------------------------------------------------------------------------------