If the intention is topic identification, trend monitoring, or sentiment analysis, I would expect a service like this to be able to drop the stopwords or at least report them separately. At the moment, by switching to aggregate data from the last sixty days, these entries are reported which I don't think should be: "all" (10.43%), "only" (6.92%), "because" (9.5%), "when" (7.88%), "even" (7.07%), "has" (10.72%), "which" (10.05%), etc.
Totally agree. Currently, the algorithm fails at times to drop stopwords which are not significant when trying to understand topics or trends. Should be able to quickly remove stopwords specifically since they are in a standard list. Thank you for the feedback.
9 comments
[ 2.8 ms ] story [ 28.0 ms ] thread