The example ApplyCustomWordEmbedding is not correct, because the method NormalizeText in ML 1.6, has by default keepPunctuation = true, which leads to extracting "product.", not "product". The output is as shown, but the explanation is not, because the 2nd position of the output should be -2, if "product" is considered. I suggest either to remove the "product" row from the custom embeddings, or modify the example accordingly.
Document Details
⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.
The example ApplyCustomWordEmbedding is not correct, because the method NormalizeText in ML 1.6, has by default keepPunctuation = true, which leads to extracting "product.", not "product". The output is as shown, but the explanation is not, because the 2nd position of the output should be -2, if "product" is considered. I suggest either to remove the "product" row from the custom embeddings, or modify the example accordingly.
Document Details
⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.