Precision and information gain of SAX* in the hashtag clustering task (one-year stream).
Golden Classifications: . | SAX* τ=0 . | SAX* τ=2 . | Baseline Clusters . | |||
---|---|---|---|---|---|---|
TSUR . | TWUBS . | TSUR . | TWUBS . | TSUR . | TWUBS . | |
. | (|Cat.|=9) . | (|Cat.|=32) . | (|Cat.|=9) . | (|Cat.|=32) . | (|Cat.|=9) . | (|Cat.|=32) . |
Average NIG | 0.967 | 0.778 | 0.98 | 0.82 | 0.77 | 0.6005 |
Stand Dev (NIG) | 0.042 | 0.1002 | 0.043 | 0.1066 | 0.25 | 0.27 |
Average Precision | 0.88 | 0.77 | 0.97 | 0.76 | 0.73 | 0.73 |
Stand Dev (Precision) | 0.127 | 0.128 | 0.085 | 0.26 | 0.27 | 0.29 |
Average # of clusters | 4.85 | 7.86 | 2.45 | 2.94 | 6.2 | 4.5 |
with |ci| > 1 in Wi |
Golden Classifications: . | SAX* τ=0 . | SAX* τ=2 . | Baseline Clusters . | |||
---|---|---|---|---|---|---|
TSUR . | TWUBS . | TSUR . | TWUBS . | TSUR . | TWUBS . | |
. | (|Cat.|=9) . | (|Cat.|=32) . | (|Cat.|=9) . | (|Cat.|=32) . | (|Cat.|=9) . | (|Cat.|=32) . |
Average NIG | 0.967 | 0.778 | 0.98 | 0.82 | 0.77 | 0.6005 |
Stand Dev (NIG) | 0.042 | 0.1002 | 0.043 | 0.1066 | 0.25 | 0.27 |
Average Precision | 0.88 | 0.77 | 0.97 | 0.76 | 0.73 | 0.73 |
Stand Dev (Precision) | 0.127 | 0.128 | 0.085 | 0.26 | 0.27 | 0.29 |
Average # of clusters | 4.85 | 7.86 | 2.45 | 2.94 | 6.2 | 4.5 |
with |ci| > 1 in Wi |