Cleanup

src/poalses/jonathan/dialect/dialect_nlp.clj

@@ -42,9 +42,6 @@
                                "ner"]
                   :quote      {:extractUnclosedQuotes "true"}}))
 
-
-(def bad-words #{"why" "cause"})
-
 ;; Word sets that will show a sentence as being of that dialect
 
 (def australian-words #{"incorrect" "why"})
@@ -73,22 +70,6 @@
     (if (empty? dialects) [:standard] dialects)))
 
 
-
-;; Another failed attempt
-;(defn detect-sentence-dialect [sentence]
-;  (let [dialects []
-;        tokens   (dl/tokens sentence)]
-;    (when (some australian-words (dl/text (dl/tokens tokens)))
-;      (let [dialects (conj dialects :australian)]
-;        (when (some scottish-words (dl/text (dl/tokens tokens)))
-;          (let [dialects (conj dialects :scottish)]
-;            (when (some american-words (dl/text (dl/tokens tokens)))
-;              (let [ dialects (conj dialects :american)]
-;                (if (empty? dialects) (conj dialects :standard))
-;                dialects))))))))
-
-
-
 ;; Take a text sample and separate it into its sentences, then for each sentence find its dialects, and return the most common dialect
 ;; A sentence can have an indeterminate number of dialects associated with it, as detect-sentence-dialects can return a collection,
 ;;when no dialect can be detected it defaults to standard. (IE if there's a sample with 3 sentences, one reads as scottish,