Short Text Clustering on SearchSnippets
85.2AccuracySCCL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SCCLAveraging=Average over five random runs2021.03 | 85.2 | 71.1 | |
| HAC-SDAlgorithm=Hierarchical Agglomerative Clustering2021.03 | 82.7 | 63.8 | |
| DECAMPComment=Proposed2021.04 | 80.6 | — | |
| RSTC2023.05 | 80.1 | 69.74 | |
| SCCL2023.05 | 79.9 | 63.78 | |
| LapDMM-WComment=Baseline (using add. info)2021.04 | 79.3 | — | |
| RSTC-Icontrastive learning=instance-wise, clustering=k-means2023.05 | 79.26 | 68.03 | |
| DECE2EComment=Proposed2021.04 | 77.8 | — | |
| DEC-SIFComment=Baseline (using add. info)2021.04 | 77.1 | — | |
| Self-TrainAlgorithm=SIF + Autoencoder2021.03 | 77.1 | 56.7 | |
| STC²-LE2017.01 | 77.09 | — | |
| STC²-LPI2017.01 | 77.01 | — | |
| STC2Comment=Baseline (using add. info)2021.04 | 77 | — | |
| STCCAlgorithm=Word2Vec + CNN + K-means2021.03 | 77 | 63.2 | |
| STC2-LPI2023.05 | 76.98 | 62.56 | |
| DECComment=Baseline2021.04 | 76.9 | — | |
| LapDMM-TComment=Baseline2021.04 | 76.1 | — | |
| STC²-LSA2017.01 | 73.09 | — | |
| Self-Train2023.05 | 72.69 | 56.74 | |
| RSTC-OTpseudo-labels=traditional OT2023.05 | 70.79 | 59.3 | |
| Para2vec2017.01 | 69.07 | — | |
| STC²-AE2017.01 | 68.34 | — | |
| RecNNPooling=Ave.2017.01 | 65.59 | — | |
| RecNNPooling=Top+Ave.2017.01 | 65.53 | — | |
| K-means_IC2023.05 | 63.84 | 42.77 | |
| RSTC-Ccontrastive learning=class-wise, pseudo-labels=SAOT2023.05 | 62.59 | 44.02 | |
| SBERT(k-means)2023.05 | 55.83 | 32.07 | |
| K-meansRepresentation=TF-IDF2017.01 | 33.77 | — | |
| SkipVecType=Combine2017.01 | 33.58 | — | |
| TF-IDFFeature dimension=1500, Algorithm=K-means2021.03 | 31.5 | 19.2 | |
| GMMComment=Baseline2021.04 | 31 | — | |
| TF-IDF2023.05 | 30.85 | 18.67 | |
| SkipVecType=Bi2017.01 | 29.24 | — | |
| SkipVecType=Uni2017.01 | 28.23 | — | |
| K-meansRepresentation=TF2017.01 | 24.75 | — | |
| BoWFeature dimension=1500, Algorithm=K-means2021.03 | 24.3 | 9.3 | |
| BOW2023.05 | 23.67 | 9 | |
| RecNNPooling=Top2017.01 | 21.21 | — | |
| DECComment=Baseline2021.04 | — | 64.9 | |
| DEC-SIFComment=Baseline (using add. info)2021.04 | — | 56.7 | |
| DECAMPComment=Proposed2021.04 | — | 69.5 | |
| DECE2EComment=Proposed2021.04 | — | 67 | |
| K-meansInput=TF2017.01 | — | 9.03 | |
| K-meansInput=TF-IDF2017.01 | — | 21.4 | |
| LapDMM-TComment=Baseline2021.04 | — | 63.4 | |
| LapDMM-WComment=Baseline (using add. info)2021.04 | — | 65.3 | |
| Para2vec2017.01 | — | 50.51 | |
| RecNNPooling=Top2017.01 | — | 4.04 | |
| RecNNPooling=Ave.2017.01 | — | 50.55 | |
| RecNNPooling=Top+Ave.2017.01 | — | 50.44 | |
| SkipVecMode=Uni2017.01 | — | 10.98 | |
| SkipVecMode=Bi2017.01 | — | 9.27 | |
| SkipVecMode=Combine2017.01 | — | 13.85 | |
| STC2Variant=AE2017.01 | — | 54.01 | |
| STC2Variant=LSA2017.01 | — | 54.53 | |
| STC2Variant=LE2017.01 | — | 63.16 | |
| STC2Variant=LPI2017.01 | — | 62.94 | |
| STC2Comment=Baseline (using add. info)2021.04 | — | 62.9 |