Update analyzer.py

analyzer.py

@@ -728,12 +728,12 @@ def analyze_sentiment_batch(
             return _simple_lexicon_sentiment(texts)
         for text in texts:
             if not text or len(text.strip()) < 3:
-                results.append({"label": "NEUTRAL", "score": 0.0, "compound": 0.0})
+                results.append({"label": "Neutral", "score": 0.0, "compound": 0.0})
                 continue
             vs = vader.polarity_scores(text)
             c  = vs["compound"]
             results.append({
-                "label":    "POSITIVELY ENGAGEMENT" if c >= 0.05 else ("NEGATIVELY ENGAGEMENT" if c <= -0.05 else "NEUTRAL"),
+                "label":    "Positively Engagement" if c >= 0.05 else ("Negatively Engagement" if c <= -0.05 else "Neutral"),
                 "score":    abs(c),
                 "compound": c,
             })
@@ -743,15 +743,17 @@ def analyze_sentiment_batch(
             chunk = texts[i: i + batch_size]
             safe  = [t[:1000] if t else " " for t in chunk]
             try:
+                _hf_label_map = {"POSITIVE": "Positively Engagement", "NEGATIVE": "Negatively Engagement"}
                 for r in pipe(safe):
+                    mapped = _hf_label_map.get(r["label"], "Neutral")
                     results.append({
-                        "label":    r["label"],
+                        "label":    mapped,
                         "score":    round(r["score"], 4),
                         "compound": r["score"] if r["label"] == "POSITIVE" else -r["score"],
                     })
             except Exception:
                 for _ in chunk:
-                    results.append({"label": "NEUTRAL", "score": 0.5, "compound": 0.0})
+                    results.append({"label": "Neutral", "score": 0.5, "compound": 0.0})
     return results
 
 
@@ -762,28 +764,28 @@ def _simple_lexicon_sentiment(texts: List[str]) -> List[Dict]:
     for text in texts:
         words = set(text.lower().split())
         p = len(words & pos); n = len(words & neg)
-        if   p > n: out.append({"label": "POSITIVELY ENGAGEMENT", "score": 0.7, "compound":  0.5})
-        elif n > p: out.append({"label": "NEGATIVE ENGAGEMENT", "score": 0.7, "compound": -0.5})
-        else:       out.append({"label": "NEUTRAL",  "score": 0.5, "compound":  0.0})
+        if   p > n: out.append({"label": "Positively Engagement", "score": 0.7, "compound":  0.5})
+        elif n > p: out.append({"label": "Negatively Engagement", "score": 0.7, "compound": -0.5})
+        else:       out.append({"label": "Neutral",               "score": 0.5, "compound":  0.0})
     return out
 
 
 def sentiment_summary(results: List[Dict]) -> Dict:
     if not results:
-        return {"POSITIVE": 0, "NEGATIVE": 0, "NEUTRAL": 0, "total": 0,
-                "avg_compound": 0.0, "pos_pct": 0, "neg_pct": 0, "neu_pct": 0}
+        return {"Positively Engagement": 0, "Negatively Engagement": 0, "Neutral": 0,
+                "total": 0, "avg_compound": 0.0, "pos_pct": 0, "neg_pct": 0, "neu_pct": 0}
     counts = Counter(r["label"] for r in results)
     total  = len(results)
     avg    = float(np.mean([r.get("compound", 0.0) for r in results]))
     return {
-        "POSITIVELY ENGAGEMENT":     counts.get("POSITIVE", 0),
-        "NEGATIVELY ENGAGEMENT":     counts.get("NEGATIVE", 0),
-        "NEUTRAL":      counts.get("NEUTRAL",  0),
+        "Positively Engagement": counts.get("Positively Engagement", 0),
+        "Negatively Engagement": counts.get("Negatively Engagement", 0),
+        "Neutral":               counts.get("Neutral",               0),
         "total":        total,
         "avg_compound": round(avg, 3),
-        "pos_pct":      round(counts.get("POSITIVE", 0) / total * 100, 1),
-        "neg_pct":      round(counts.get("NEGATIVE", 0) / total * 100, 1),
-        "neu_pct":      round(counts.get("NEUTRAL",  0) / total * 100, 1),
+        "pos_pct":      round(counts.get("Positively Engagement", 0) / total * 100, 1),
+        "neg_pct":      round(counts.get("Negatively Engagement", 0) / total * 100, 1),
+        "neu_pct":      round(counts.get("Neutral",               0) / total * 100, 1),
     }
 
 
@@ -822,6 +824,6 @@ def sentiment_weighted_keywords(
     for text, sent in zip(texts, sentiment_results):
         tokens = [t for t in re.findall(r"[a-zA-Z]{3,}", text.lower()) if t not in STOPWORDS]
         weight = sent.get("score", 0.5)
-        if   sent["label"] == "POSITIVELY ENGAGEMENT": pos_freq.update({t: weight for t in tokens})
-        elif sent["label"] == "NEGATIVELY ENGAGEMENT": neg_freq.update({t: weight for t in tokens})
+        if   sent["label"] == "Positively Engagement": pos_freq.update({t: weight for t in tokens})
+        elif sent["label"] == "Negatively Engagement": neg_freq.update({t: weight for t in tokens})
     return pos_freq.most_common(top_n), neg_freq.most_common(top_n)