Text analytics -stemming vs lemmatization



Can someone give me 2 business cases with examples where stemming is better than lemmatisation?


If stemming is preferred, it is usually for the below two reasons

  1. Processing of huge volume of corpus where either speed is vital or POS cannot be used
  2. Web or app query where the search is tolerant to word inflections not for the meaning of query (ex: search the corpus for the word and its inflections and not required to match different words having similar meanings)


Thanks Prakash.