Create question_classifier.py

question_classifier.py

@@ -0,0 +1,66 @@
+import re
+
+
+def classify_question(question_text: str, category: str) -> dict:
+    q = question_text.lower()
+
+    if category == "Quantitative":
+
+        if "percent" in q or "%" in q:
+            if "then" in q:
+                return {"topic": "percent", "type": "successive_percent"}
+            return {"topic": "percent", "type": "percent_change"}
+
+        if "ratio" in q:
+            return {"topic": "ratio", "type": "ratio_total"}
+
+        if "probability" in q or "chosen at random" in q:
+            return {"topic": "probability", "type": "simple_probability"}
+
+        if "divisible" in q or "remainder" in q:
+            return {"topic": "number_theory", "type": "divisibility"}
+
+        if "|" in q:
+            return {"topic": "algebra", "type": "absolute_value"}
+
+        if "circle" in q or "radius" in q or "circumference" in q:
+            return {"topic": "geometry", "type": "circle_geometry"}
+
+        if "average" in q or "mean" in q:
+            return {"topic": "statistics", "type": "average"}
+
+        if "sequence" in q:
+            return {"topic": "sequence", "type": "geometric_sequence"}
+
+        if "=" in q:
+            return {"topic": "algebra", "type": "equation"}
+
+        return {"topic": "quant", "type": "general"}
+
+    if category == "DataInsight":
+
+        if "percent" in q:
+            return {"topic": "percent", "type": "percent_change"}
+
+        if "mean" in q or "median" in q:
+            return {"topic": "statistics", "type": "distribution"}
+
+        if "correlation" in q or "scatter" in q:
+            return {"topic": "statistics", "type": "correlation"}
+
+        return {"topic": "data", "type": "general"}
+
+    if category == "Verbal":
+
+        if "meaning" in q:
+            return {"topic": "vocabulary", "type": "definition"}
+
+        if "grammatically" in q:
+            return {"topic": "grammar", "type": "sentence_correction"}
+
+        if "argument" in q or "author" in q:
+            return {"topic": "reasoning", "type": "argument_analysis"}
+
+        return {"topic": "verbal", "type": "general"}
+
+    return {"topic": "unknown", "type": "unknown"}