Soundex is not enough. We went through metaphone and double-metaphone as well. The last showed the best performance when combined with simple ways to reduce the search space (e.g., names that start with the same alphabet).
But it still had too many false positives and negatives. We ended up using a much simpler approach of manually labeling Top N most frequent names. On Tue, Aug 14, 2018 at 7:58 AM, Pradeep Bhatt <[email protected]> wrote: > Hi All, > > What is the best way to know if two words are phonetically similar > > e.g *Some similar *words > > Pradeep - Pradip > Thakkkar - Thakkar > Rathod - Rathor > Swetha - Sweta > bhen - ben > Sumandev - Sumandeb > > *Non - Similar* > Ramesh - Rajesh > > This is needed for spelling mistakes introduced when translating from > indian languages to English. > > Does Soundex work well for Indian names ? > > Regards, > Pradeep > > > > -- > Datameet is a community of Data Science enthusiasts in India. Know more > about us by visiting http://datameet.org > --- > You received this message because you are subscribed to the Google Groups > "datameet" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org --- You received this message because you are subscribed to the Google Groups "datameet" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
