The necessary packages can be found in requirements.txt
After preparing the environment, unpackage data.zip, and just run bash auto.sh to obtain the results.
Some important parameters: --pb denotes whether to use progressive blocking --agg denotes which aggregation method is used. "arith" represents the arithmetic mean and "harm" represents the harmonic mean. --folder denotes the dataset
Due to the instability of embedding-based methods, it is acceptable that the results fluctuate a little bit (±1%) when running code repeatedly.