feat: add descriptions to topics to improve accuracy

Also upgraded to more powerful model
This commit is contained in:
2026-02-08 15:10:11 +00:00
parent b019885b2f
commit a9d63c7041
4 changed files with 27 additions and 13 deletions

View File

@@ -42,7 +42,7 @@ class StatGen:
df["weekday"] = df["dt"].dt.day_name()
add_emotion_cols(df, "content")
add_topic_col(df, "content", self.domain_topics)
add_topic_col(df, "title", "content", self.domain_topics)
def _tokenize(self, text: str):
tokens = re.findall(r"\b[a-z]{3,}\b", text)