Performance of systems (per-tag and overall precision (P), recall (R) and F value (F)) and statistical significance tests
. | Performance of systems (10-fold cross-validation) . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
MIST1 . | MCRF1 . | MIST2 . | MCRF2 . | |||||||||
P . | R . | F . | P . | R . | F . | P . | R . | F . | P . | R . | F . | |
AGE | 94.47 | 92.31 | 93.38 | 96.7 | 90.42 | 93.45 | 95.87 | 92.46 | 94.13 | 96.69 | 90 | 93.22 |
DATE | 95.77 | 97.12 | 96.44 | 97.97 | 96.61 | 97.29 | 97.25 | 97.76 | 97.5 | 97.95 | 96.98 | 97.46 |
ID | 90.58 | 92.64 | 91.6 | 97.23 | 95.64 | 96.43 | 91.17 | 93.38 | 92.26 | 97.27 | 95.7 | 96.48 |
INST | 90.59 | 86.41 | 88.45 | 93.18 | 85.01 | 88.91 | 90.61 | 87.06 | 88.8 | 93.25 | 85.26 | 89.08 |
LOC | 79.82 | 67.93 | 73.4 | 86.12 | 68.94 | 76.58 | 78.92 | 69.95 | 74.16 | 87.38 | 69.95 | 77.7 |
NAME | 93.19 | 88.64 | 90.86 | 95.62 | 86.99 | 91.1 | 92.48 | 94.16 | 93.31 | 94.47 | 94.56 | 94.52 |
OTH | 77.21 | 77.39 | 77.3 | 83.94 | 74.17 | 78.76 | 78.17 | 77.13 | 77.65 | 84.68 | 73.87 | 78.91 |
PH | 90.06 | 93.04 | 91.52 | 94.08 | 90.64 | 92.33 | 91.44 | 93.95 | 92.68 | 94.42 | 90.87 | 92.61 |
All | 92.05 | 91.02 | 91.54 | 95.25 | 89.86 | 92.48 | 92.79 | 92.81 | 92.8 | 95.08 | 91.92 | 93.48 |
. | Performance of systems (10-fold cross-validation) . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
MIST1 . | MCRF1 . | MIST2 . | MCRF2 . | |||||||||
P . | R . | F . | P . | R . | F . | P . | R . | F . | P . | R . | F . | |
AGE | 94.47 | 92.31 | 93.38 | 96.7 | 90.42 | 93.45 | 95.87 | 92.46 | 94.13 | 96.69 | 90 | 93.22 |
DATE | 95.77 | 97.12 | 96.44 | 97.97 | 96.61 | 97.29 | 97.25 | 97.76 | 97.5 | 97.95 | 96.98 | 97.46 |
ID | 90.58 | 92.64 | 91.6 | 97.23 | 95.64 | 96.43 | 91.17 | 93.38 | 92.26 | 97.27 | 95.7 | 96.48 |
INST | 90.59 | 86.41 | 88.45 | 93.18 | 85.01 | 88.91 | 90.61 | 87.06 | 88.8 | 93.25 | 85.26 | 89.08 |
LOC | 79.82 | 67.93 | 73.4 | 86.12 | 68.94 | 76.58 | 78.92 | 69.95 | 74.16 | 87.38 | 69.95 | 77.7 |
NAME | 93.19 | 88.64 | 90.86 | 95.62 | 86.99 | 91.1 | 92.48 | 94.16 | 93.31 | 94.47 | 94.56 | 94.52 |
OTH | 77.21 | 77.39 | 77.3 | 83.94 | 74.17 | 78.76 | 78.17 | 77.13 | 77.65 | 84.68 | 73.87 | 78.91 |
PH | 90.06 | 93.04 | 91.52 | 94.08 | 90.64 | 92.33 | 91.44 | 93.95 | 92.68 | 94.42 | 90.87 | 92.61 |
All | 92.05 | 91.02 | 91.54 | 95.25 | 89.86 | 92.48 | 92.79 | 92.81 | 92.8 | 95.08 | 91.92 | 93.48 |
. | Token-level performance for best system (MCRF2) . | |||||
---|---|---|---|---|---|---|
Token-level . | Token-level + tag-blind . | |||||
P . | R . | F . | P . | R . | F . | |
AGE | 98.21 | 93.40 | 95.75 | 93.42 | ||
DATE | 98.17 | 96.87 | 97.52 | 96.98 | ||
ID | 97.57 | 95.49 | 96.52 | 96.43 | ||
INST | 97.48 | 92.74 | 95.05 | 94.79 | ||
LOC | 97.95 | 93.92 | 95.89 | 96.03 | ||
NAME | 97.26 | 97.38 | 97.32 | 97.53 | ||
OTH | 86.81 | 76.45 | 81.30 | 78.31 | ||
PH | 97.13 | 93.40 | 95.23 | 94.85 | ||
All | 96.68 | 93.77 | 95.20 | 97.42 | 94.49 | 95.93 |
. | Token-level performance for best system (MCRF2) . | |||||
---|---|---|---|---|---|---|
Token-level . | Token-level + tag-blind . | |||||
P . | R . | F . | P . | R . | F . | |
AGE | 98.21 | 93.40 | 95.75 | 93.42 | ||
DATE | 98.17 | 96.87 | 97.52 | 96.98 | ||
ID | 97.57 | 95.49 | 96.52 | 96.43 | ||
INST | 97.48 | 92.74 | 95.05 | 94.79 | ||
LOC | 97.95 | 93.92 | 95.89 | 96.03 | ||
NAME | 97.26 | 97.38 | 97.32 | 97.53 | ||
OTH | 86.81 | 76.45 | 81.30 | 78.31 | ||
PH | 97.13 | 93.40 | 95.23 | 94.85 | ||
All | 96.68 | 93.77 | 95.20 | 97.42 | 94.49 | 95.93 |
. | Statistical significance tests between F values obtained by systems (cross-validation evaluation) . | ||||
---|---|---|---|---|---|
MCRF1 vs MIST1 . | MCRF2 vs MIST2 . | MIST1 vs MIST2 . | MCRF2 vs MCRF1 . | MCRF2 vs gold standard . | |
p Value . | p Value . | p Value . | p Value . | p Value . | |
AGE | 0.8490 | *0.0087 | *0.0389 | 0.4922 | *0.0001 |
DATE | *0.0001 | 0.7650 | *0.0001 | 0.2470 | *0.0001 |
ID | *0.0001 | *0.0001 | 0.0996 | 0.8856 | *0.0001 |
INST | 0.3572 | 0.6087 | 0.2738 | 0.6623 | *0.0001 |
LOC | 0.0777 | 0.0553 | 0.5897 | 0.3812 | *0.0001 |
NAME | 0.4897 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
OTH | *0.0071 | *0.0180 | 0.3676 | 0.6118 | *0.0001 |
PH | 0.2936 | 0.9248 | *0.0458 | 0.6208 | *0.0001 |
All | *0.0001 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
. | Statistical significance tests between F values obtained by systems (cross-validation evaluation) . | ||||
---|---|---|---|---|---|
MCRF1 vs MIST1 . | MCRF2 vs MIST2 . | MIST1 vs MIST2 . | MCRF2 vs MCRF1 . | MCRF2 vs gold standard . | |
p Value . | p Value . | p Value . | p Value . | p Value . | |
AGE | 0.8490 | *0.0087 | *0.0389 | 0.4922 | *0.0001 |
DATE | *0.0001 | 0.7650 | *0.0001 | 0.2470 | *0.0001 |
ID | *0.0001 | *0.0001 | 0.0996 | 0.8856 | *0.0001 |
INST | 0.3572 | 0.6087 | 0.2738 | 0.6623 | *0.0001 |
LOC | 0.0777 | 0.0553 | 0.5897 | 0.3812 | *0.0001 |
NAME | 0.4897 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
OTH | *0.0071 | *0.0180 | 0.3676 | 0.6118 | *0.0001 |
PH | 0.2936 | 0.9248 | *0.0458 | 0.6208 | *0.0001 |
All | *0.0001 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
*Indicates statistical significance (p<0.05).
INST, institution; LOC, location; OTH, other; PH, phone.
Performance of systems (per-tag and overall precision (P), recall (R) and F value (F)) and statistical significance tests
. | Performance of systems (10-fold cross-validation) . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
MIST1 . | MCRF1 . | MIST2 . | MCRF2 . | |||||||||
P . | R . | F . | P . | R . | F . | P . | R . | F . | P . | R . | F . | |
AGE | 94.47 | 92.31 | 93.38 | 96.7 | 90.42 | 93.45 | 95.87 | 92.46 | 94.13 | 96.69 | 90 | 93.22 |
DATE | 95.77 | 97.12 | 96.44 | 97.97 | 96.61 | 97.29 | 97.25 | 97.76 | 97.5 | 97.95 | 96.98 | 97.46 |
ID | 90.58 | 92.64 | 91.6 | 97.23 | 95.64 | 96.43 | 91.17 | 93.38 | 92.26 | 97.27 | 95.7 | 96.48 |
INST | 90.59 | 86.41 | 88.45 | 93.18 | 85.01 | 88.91 | 90.61 | 87.06 | 88.8 | 93.25 | 85.26 | 89.08 |
LOC | 79.82 | 67.93 | 73.4 | 86.12 | 68.94 | 76.58 | 78.92 | 69.95 | 74.16 | 87.38 | 69.95 | 77.7 |
NAME | 93.19 | 88.64 | 90.86 | 95.62 | 86.99 | 91.1 | 92.48 | 94.16 | 93.31 | 94.47 | 94.56 | 94.52 |
OTH | 77.21 | 77.39 | 77.3 | 83.94 | 74.17 | 78.76 | 78.17 | 77.13 | 77.65 | 84.68 | 73.87 | 78.91 |
PH | 90.06 | 93.04 | 91.52 | 94.08 | 90.64 | 92.33 | 91.44 | 93.95 | 92.68 | 94.42 | 90.87 | 92.61 |
All | 92.05 | 91.02 | 91.54 | 95.25 | 89.86 | 92.48 | 92.79 | 92.81 | 92.8 | 95.08 | 91.92 | 93.48 |
. | Performance of systems (10-fold cross-validation) . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
MIST1 . | MCRF1 . | MIST2 . | MCRF2 . | |||||||||
P . | R . | F . | P . | R . | F . | P . | R . | F . | P . | R . | F . | |
AGE | 94.47 | 92.31 | 93.38 | 96.7 | 90.42 | 93.45 | 95.87 | 92.46 | 94.13 | 96.69 | 90 | 93.22 |
DATE | 95.77 | 97.12 | 96.44 | 97.97 | 96.61 | 97.29 | 97.25 | 97.76 | 97.5 | 97.95 | 96.98 | 97.46 |
ID | 90.58 | 92.64 | 91.6 | 97.23 | 95.64 | 96.43 | 91.17 | 93.38 | 92.26 | 97.27 | 95.7 | 96.48 |
INST | 90.59 | 86.41 | 88.45 | 93.18 | 85.01 | 88.91 | 90.61 | 87.06 | 88.8 | 93.25 | 85.26 | 89.08 |
LOC | 79.82 | 67.93 | 73.4 | 86.12 | 68.94 | 76.58 | 78.92 | 69.95 | 74.16 | 87.38 | 69.95 | 77.7 |
NAME | 93.19 | 88.64 | 90.86 | 95.62 | 86.99 | 91.1 | 92.48 | 94.16 | 93.31 | 94.47 | 94.56 | 94.52 |
OTH | 77.21 | 77.39 | 77.3 | 83.94 | 74.17 | 78.76 | 78.17 | 77.13 | 77.65 | 84.68 | 73.87 | 78.91 |
PH | 90.06 | 93.04 | 91.52 | 94.08 | 90.64 | 92.33 | 91.44 | 93.95 | 92.68 | 94.42 | 90.87 | 92.61 |
All | 92.05 | 91.02 | 91.54 | 95.25 | 89.86 | 92.48 | 92.79 | 92.81 | 92.8 | 95.08 | 91.92 | 93.48 |
. | Token-level performance for best system (MCRF2) . | |||||
---|---|---|---|---|---|---|
Token-level . | Token-level + tag-blind . | |||||
P . | R . | F . | P . | R . | F . | |
AGE | 98.21 | 93.40 | 95.75 | 93.42 | ||
DATE | 98.17 | 96.87 | 97.52 | 96.98 | ||
ID | 97.57 | 95.49 | 96.52 | 96.43 | ||
INST | 97.48 | 92.74 | 95.05 | 94.79 | ||
LOC | 97.95 | 93.92 | 95.89 | 96.03 | ||
NAME | 97.26 | 97.38 | 97.32 | 97.53 | ||
OTH | 86.81 | 76.45 | 81.30 | 78.31 | ||
PH | 97.13 | 93.40 | 95.23 | 94.85 | ||
All | 96.68 | 93.77 | 95.20 | 97.42 | 94.49 | 95.93 |
. | Token-level performance for best system (MCRF2) . | |||||
---|---|---|---|---|---|---|
Token-level . | Token-level + tag-blind . | |||||
P . | R . | F . | P . | R . | F . | |
AGE | 98.21 | 93.40 | 95.75 | 93.42 | ||
DATE | 98.17 | 96.87 | 97.52 | 96.98 | ||
ID | 97.57 | 95.49 | 96.52 | 96.43 | ||
INST | 97.48 | 92.74 | 95.05 | 94.79 | ||
LOC | 97.95 | 93.92 | 95.89 | 96.03 | ||
NAME | 97.26 | 97.38 | 97.32 | 97.53 | ||
OTH | 86.81 | 76.45 | 81.30 | 78.31 | ||
PH | 97.13 | 93.40 | 95.23 | 94.85 | ||
All | 96.68 | 93.77 | 95.20 | 97.42 | 94.49 | 95.93 |
. | Statistical significance tests between F values obtained by systems (cross-validation evaluation) . | ||||
---|---|---|---|---|---|
MCRF1 vs MIST1 . | MCRF2 vs MIST2 . | MIST1 vs MIST2 . | MCRF2 vs MCRF1 . | MCRF2 vs gold standard . | |
p Value . | p Value . | p Value . | p Value . | p Value . | |
AGE | 0.8490 | *0.0087 | *0.0389 | 0.4922 | *0.0001 |
DATE | *0.0001 | 0.7650 | *0.0001 | 0.2470 | *0.0001 |
ID | *0.0001 | *0.0001 | 0.0996 | 0.8856 | *0.0001 |
INST | 0.3572 | 0.6087 | 0.2738 | 0.6623 | *0.0001 |
LOC | 0.0777 | 0.0553 | 0.5897 | 0.3812 | *0.0001 |
NAME | 0.4897 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
OTH | *0.0071 | *0.0180 | 0.3676 | 0.6118 | *0.0001 |
PH | 0.2936 | 0.9248 | *0.0458 | 0.6208 | *0.0001 |
All | *0.0001 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
. | Statistical significance tests between F values obtained by systems (cross-validation evaluation) . | ||||
---|---|---|---|---|---|
MCRF1 vs MIST1 . | MCRF2 vs MIST2 . | MIST1 vs MIST2 . | MCRF2 vs MCRF1 . | MCRF2 vs gold standard . | |
p Value . | p Value . | p Value . | p Value . | p Value . | |
AGE | 0.8490 | *0.0087 | *0.0389 | 0.4922 | *0.0001 |
DATE | *0.0001 | 0.7650 | *0.0001 | 0.2470 | *0.0001 |
ID | *0.0001 | *0.0001 | 0.0996 | 0.8856 | *0.0001 |
INST | 0.3572 | 0.6087 | 0.2738 | 0.6623 | *0.0001 |
LOC | 0.0777 | 0.0553 | 0.5897 | 0.3812 | *0.0001 |
NAME | 0.4897 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
OTH | *0.0071 | *0.0180 | 0.3676 | 0.6118 | *0.0001 |
PH | 0.2936 | 0.9248 | *0.0458 | 0.6208 | *0.0001 |
All | *0.0001 | *0.0001 | *0.0001 | *0.0001 | *0.0001 |
*Indicates statistical significance (p<0.05).
INST, institution; LOC, location; OTH, other; PH, phone.
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.