Hi all, I'm a new user of SAS Visutal Text Analytics and I have some issues creating new category rules. My desire is to identify all the documents in which two concepts are present in a same part of a sentence, delimited by semicolons (or by other elements). Here a simple example of my situation and what I'd like to identify. My concepts are: - VEICHLES -> CLASSIFIER: CAR - TYPES-> CLASSIFIER: SPORT Documents are: 1. other sentences. I drive a sport car. other sentences. 2. other sentences. I play a sport; I usually drive a car. other sentences. 3. other sentences. I drive a car; I play a sport. other sentences. My category must contain just the documents like the 1, identifying "sport car" in a part of the same sentence. The two concepts must belong to the same sentence and they must not be separated by ";". I need to generalize this behaviour considering ":" , "-" and other marks. I also try to use a predicate_rule like PREDICATE_RULE: (veichles, type):(UNLESS, ";", (SENT, "_veichles{VEHICLES}","_sport{TYPES}")) but I can't figure out how to parametrize ";" argument (the same behaviour must occur with ":" too). Do anyone have solutions for this situation? Thank in advance Giorgio
... View more