Skip to content

MECOL2 LEX

Vocabulary Knowledge (MECOL2)

Test

ModelUnseen Reader RMSEUnseen Text RMSEUnseen Text and Reader RMSEAverage RMSEUnseen Reader MAEUnseen Text MAEUnseen Text and Reader MAEAverage MAEUnseen Reader R²Unseen Text R²Unseen Text and Reader R²Average R²
Majority Class / Chance12.47 ± 0.112.46 ± 0.012.39 ± 0.112.45 ± 0.110.41 ± 0.110.4 ± 0.010.32 ± 0.110.4 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
Reading Speed28.98 ± 14.328.92 ± 14.328.97 ± 14.328.95 ± 14.327.18 ± 14.527.14 ± 14.527.14 ± 14.627.15 ± 14.5-9.86 ± 8.5-9.58 ± 8.3-10.06 ± 8.7-9.76 ± 8.5
Text-Only Roberta12.47 ± 0.112.46 ± 0.012.39 ± 0.112.45 ± 0.110.41 ± 0.110.41 ± 0.010.32 ± 0.110.4 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
Logistic Regression [meziere2023using]10.37 ± 0.110.31 ± 0.110.32 ± 0.210.34 ± 0.08.35 ± 0.18.31 ± 0.18.31 ± 0.28.32 ± 0.00.31 ± 0.00.31 ± 0.00.3 ± 0.00.31 ± 0.0
SVM [hollenstein2023zuco]10.98 ± 0.110.94 ± 0.110.84 ± 0.110.94 ± 0.08.78 ± 0.08.76 ± 0.08.68 ± 0.18.76 ± 0.00.22 ± 0.00.23 ± 0.00.23 ± 0.00.23 ± 0.0
Random Forest [makowski2024detection]10.2 ± 0.19.93 ± 0.010.24 ± 0.210.09 ± 0.08.2 ± 0.17.95 ± 0.08.26 ± 0.28.1 ± 0.00.33 ± 0.00.36 ± 0.00.31 ± 0.00.34 ± 0.0
AhnRNN [ahn2020towards]12.48 ± 0.112.46 ± 0.012.4 ± 0.212.46 ± 0.110.42 ± 0.110.4 ± 0.010.32 ± 0.110.4 ± 0.1-0.0 ± 0.00.0 ± 0.0-0.01 ± 0.0-0.0 ± 0.0
AhnCNN [ahn2020towards]12.28 ± 0.112.3 ± 0.012.2 ± 0.112.27 ± 0.010.25 ± 0.110.27 ± 0.010.16 ± 0.110.24 ± 0.00.03 ± 0.00.03 ± 0.00.03 ± 0.00.03 ± 0.0
BEyeLSTM [reich_inferring_2022]12.65 ± 0.312.42 ± 0.113.42 ± 0.912.68 ± 0.310.38 ± 0.210.33 ± 0.110.41 ± 0.310.36 ± 0.1-0.03 ± 0.00.01 ± 0.0-0.19 ± 0.1-0.04 ± 0.0
PLM-AS [Yang2023PLMASPL]12.48 ± 0.112.46 ± 0.012.4 ± 0.212.46 ± 0.110.42 ± 0.110.41 ± 0.010.32 ± 0.110.4 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.01 ± 0.0-0.0 ± 0.0
PLM-AS-RM [haller2022eye]31.8 ± 0.331.81 ± 0.131.69 ± 0.431.79 ± 0.229.28 ± 0.329.28 ± 0.229.2 ± 0.429.27 ± 0.2-5.53 ± 0.1-5.52 ± 0.1-5.57 ± 0.2-5.52 ± 0.1
RoBERTEye-W [Shubi2024Finegrained]11.11 ± 0.211.2 ± 0.111.16 ± 0.211.16 ± 0.19.15 ± 0.19.21 ± 0.19.2 ± 0.29.18 ± 0.00.2 ± 0.00.19 ± 0.00.19 ± 0.00.2 ± 0.0
RoBERTEye-F [Shubi2024Finegrained]11.28 ± 0.311.18 ± 0.211.32 ± 0.311.25 ± 0.29.28 ± 0.29.19 ± 0.19.32 ± 0.39.25 ± 0.10.18 ± 0.00.19 ± 0.00.16 ± 0.00.18 ± 0.0
MAG-Eye [Shubi2024Finegrained]12.47 ± 0.112.46 ± 0.012.39 ± 0.112.46 ± 0.110.42 ± 0.110.41 ± 0.010.32 ± 0.110.4 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
PostFusion-Eye [Shubi2024Finegrained]13.44 ± 0.413.99 ± 0.815.26 ± 1.214.0 ± 0.410.63 ± 0.311.17 ± 0.711.92 ± 1.011.04 ± 0.4-0.17 ± 0.0-0.28 ± 0.1-0.54 ± 0.2-0.27 ± 0.1

Validation

ModelUnseen Reader RMSEUnseen Text RMSEUnseen Text and Reader RMSEAverage RMSEUnseen Reader MAEUnseen Text MAEUnseen Text and Reader MAEAverage MAEUnseen Reader R²Unseen Text R²Unseen Text and Reader R²Average R²
Majority Class / Chance12.49 ± 0.112.45 ± 0.112.39 ± 0.112.45 ± 0.110.42 ± 0.110.4 ± 0.110.31 ± 0.110.39 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
Reading Speed28.76 ± 14.229.0 ± 14.428.61 ± 14.228.82 ± 14.226.97 ± 14.427.22 ± 14.626.83 ± 14.427.04 ± 14.5-9.09 ± 7.9-9.63 ± 8.3-9.07 ± 7.9-9.28 ± 8.0
Text-Only Roberta12.49 ± 0.112.45 ± 0.112.38 ± 0.112.45 ± 0.110.42 ± 0.110.4 ± 0.110.31 ± 0.110.39 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
Logistic Regression [meziere2023using]10.37 ± 0.210.27 ± 0.110.32 ± 0.110.32 ± 0.18.35 ± 0.18.29 ± 0.18.31 ± 0.18.32 ± 0.10.31 ± 0.00.32 ± 0.00.3 ± 0.00.31 ± 0.0
SVM [hollenstein2023zuco]10.95 ± 0.110.9 ± 0.010.79 ± 0.110.9 ± 0.18.77 ± 0.18.74 ± 0.08.63 ± 0.18.73 ± 0.00.23 ± 0.00.23 ± 0.00.24 ± 0.00.23 ± 0.0
Random Forest [makowski2024detection]10.19 ± 0.19.71 ± 0.010.26 ± 0.110.01 ± 0.18.16 ± 0.17.79 ± 0.08.32 ± 0.28.04 ± 0.00.33 ± 0.00.39 ± 0.00.31 ± 0.00.35 ± 0.0
AhnRNN [ahn2020towards]12.48 ± 0.112.45 ± 0.112.37 ± 0.112.45 ± 0.110.41 ± 0.110.41 ± 0.110.3 ± 0.110.39 ± 0.10.0 ± 0.0-0.0 ± 0.00.0 ± 0.00.0 ± 0.0
AhnCNN [ahn2020towards]12.33 ± 0.112.27 ± 0.112.21 ± 0.112.28 ± 0.110.28 ± 0.110.25 ± 0.110.17 ± 0.210.25 ± 0.10.02 ± 0.00.03 ± 0.00.02 ± 0.00.03 ± 0.0
BEyeLSTM [reich_inferring_2022]12.64 ± 0.412.33 ± 0.112.55 ± 0.412.51 ± 0.310.4 ± 0.210.3 ± 0.110.29 ± 0.210.34 ± 0.1-0.03 ± 0.00.02 ± 0.0-0.03 ± 0.1-0.01 ± 0.0
PLM-AS [Yang2023PLMASPL]12.48 ± 0.112.45 ± 0.112.37 ± 0.112.45 ± 0.110.41 ± 0.110.41 ± 0.110.3 ± 0.110.39 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
PLM-AS-RM [haller2022eye]31.81 ± 0.331.78 ± 0.031.7 ± 0.431.78 ± 0.229.27 ± 0.329.25 ± 0.129.2 ± 0.429.25 ± 0.2-5.5 ± 0.2-5.52 ± 0.1-5.58 ± 0.2-5.52 ± 0.1
RoBERTEye-W [Shubi2024Finegrained]11.05 ± 0.111.11 ± 0.111.16 ± 0.111.1 ± 0.09.1 ± 0.09.11 ± 0.19.22 ± 0.29.13 ± 0.00.22 ± 0.00.2 ± 0.00.18 ± 0.00.21 ± 0.0
RoBERTEye-F [Shubi2024Finegrained]11.17 ± 0.110.98 ± 0.011.03 ± 0.111.07 ± 0.09.2 ± 0.19.04 ± 0.09.11 ± 0.19.12 ± 0.00.2 ± 0.00.22 ± 0.00.2 ± 0.00.21 ± 0.0
MAG-Eye [Shubi2024Finegrained]12.49 ± 0.112.45 ± 0.112.38 ± 0.112.45 ± 0.110.42 ± 0.110.4 ± 0.110.31 ± 0.110.39 ± 0.1-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0-0.0 ± 0.0
PostFusion-Eye [Shubi2024Finegrained]13.01 ± 0.413.72 ± 0.913.15 ± 0.313.35 ± 0.510.37 ± 0.311.0 ± 0.810.2 ± 0.110.59 ± 0.4-0.09 ± 0.1-0.23 ± 0.2-0.13 ± 0.1-0.16 ± 0.1