The text classifiers in big data analytics is very useful for identifying words that are of interest to you, such as keywords or phrases (or even entire sentences).
It's not always perfect but it does make searching through large amounts easier than using traditional methods like grep/awk with regex patterns if your looking at thousands upon tens-of -thousands rows worths per file! I dislike how difficult this tool can be when trying out new features / models since there seems no way around…