Kanium-PUBLIC/KaniumCogs
5
0
Fork 0
Code Issues 6 Pull Requests Packages Projects 1 Releases Wiki Activity

trying to fix topic extraction

Browse Source
This commit is contained in:
AllfatherHatt 2025-02-24 18:46:19 +01:00
parent 541d42c558
commit 05e1842000
1 changed files with 7 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
reginaldCog/reginald.py
View File

@ -283,17 +283,19 @@ class ReginaldCog(commands.Cog):
def extract_topics_from_summary(self, summary): def extract_topics_from_summary(self, summary):
"""Dynamically extracts the most important topics from a summary.""" """Dynamically extracts the most important topics from a summary."""
# 🔹 Extract all words from summary if isinstance(summary, dict): # ✅ Extract summary content correctly
summary = summary.get("summary", "")
if not isinstance(summary, str): # ✅ Additional safeguard
return []
keywords = re.findall(r"\b\w+\b", summary.lower()) keywords = re.findall(r"\b\w+\b", summary.lower())
# 🔹 Count word occurrences
word_counts = Counter(keywords) word_counts = Counter(keywords)
# 🔹 Remove unimportant words (common filler words) stop_words = {"the", "and", "of", "in", "to", "is", "on", "for", "with", "at", "by", "it", "this", "that"}
stop_words = {"the", "and", "of", "in", "to", "is", "on", "for", "with", "at", "by", "it", "this", "that", "his", "her"}
filtered_words = {word: count for word, count in word_counts.items() if word not in stop_words and len(word) > 2} filtered_words = {word: count for word, count in word_counts.items() if word not in stop_words and len(word) > 2}
# 🔹 Take the 5 most frequently used words as "topics"
topics = sorted(filtered_words, key=filtered_words.get, reverse=True)[:5] topics = sorted(filtered_words, key=filtered_words.get, reverse=True)[:5]
return topics return topics