Clustering on StackExchange
44ARIAgentic clustering
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Agentic clusteringk setting=Given k, predicted number of clusters (ˆk)=121, total cost ($)=$100 + $222026.05 | 44 | 73 | 56 | |
| Agentic clusteringk setting=Discover k, predicted number of clusters (ˆk)=130, total cost ($)=$100 + $232026.05 | 42 | 73 | 56 | |
| LLM-embedding+k-meansk setting=Given k, predicted number of clusters (ˆk)=121, total cost ($)=$0.392026.05 | 34 | 67 | 47 | |
| ClusterLLMk setting=Given k, predicted number of clusters (ˆk)=121, total cost ($)=$1.532026.05 | 31 | 66 | 45 | |
| BERTopick setting=Given k, predicted number of clusters (ˆk)=70, total cost ($)=$02026.05 | 17 | 51 | 30 | |
| BERTopick setting=Discover k, predicted number of clusters (ˆk)=70, total cost ($)=$02026.05 | 17 | 51 | 30 | |
| SBERT+k-meansk setting=Given k, predicted number of clusters (ˆk)=121, total cost ($)=$02026.05 | 16 | 52 | 29 | |
| TopicGPTk setting=Discover k, predicted number of clusters (ˆk)=30, total cost ($)=$100 + $662026.05 | 13 | 54 | 28 | |
| Huang & Hek setting=Discover k, predicted number of clusters (ˆk)=2764, total cost ($)=$100 + $442026.05 | 2 | 69 | 10 | |
| LDAk setting=Given k, predicted number of clusters (ˆk)=121, total cost ($)=$02026.05 | 1 | 31 | 8 |