There are valid use cases to retain the id field in the original retrieved documents in the documents that are reranked by VertexRank. This is especially important for use cases where traceability of documents throughout the chain is important.
Current assigned id based on an enumerated incrementing index works for cases where ids do not need to be tracked but isn't universally applicable.
There are valid use cases to retain the
id
field in the original retrieved documents in the documents that are reranked byVertexRank
. This is especially important for use cases where traceability of documents throughout the chain is important.Current assigned
id
based on an enumerated incrementing index works for cases whereid
s do not need to be tracked but isn't universally applicable.