evaluate pipeline for datasets

[Aisuko]

We need to know how to define the high quality datasets which means that we need a pipeline to evaluate the similarity between the questions and the datasets

https://www.kaggle.com/code/aisuko/semantic-search
https://www.kaggle.com/code/aisuko/kmeans-with-sk-learn

For small paragraph

https://www.kaggle.com/code/aisuko/in-document-search-cross-encoder