Customer Feedback Analysis Project in Python

"""Theme and rating analysis for a small synthetic feedback table."""

from __future__ import annotations

import re
import pandas as pd

THEMES = {"support": {"support", "agent", "reply"}, "usability": {"easy", "confusing", "screen", "navigation"}, "performance": {"fast", "slow", "loading"}, "content": {"lesson", "example", "course", "explanation"}}


def demo_feedback() -> pd.DataFrame:
    return pd.DataFrame({"rating": [5, 2, 4, 1, 5, 3, 4, 2, 5, 3], "comment": ["Support agent gave a fast reply", "The screen is confusing", "Clear lesson and helpful example", "Loading is very slow", "The course explanation is excellent", "Navigation is okay", "Support solved my issue", "Examples need more detail", "Easy lessons and fast pages", "Agent reply arrived later"]})


def detect_themes(text: str) -> list[str]:
    tokens = set(re.findall(r"[a-z]+", text.lower()))
    return [theme for theme, words in THEMES.items() if tokens & words] or ["other"]


def analyse_feedback(data: pd.DataFrame) -> tuple[dict[str, float], pd.DataFrame]:
    feedback = data.copy()
    feedback["theme"] = feedback["comment"].map(detect_themes)
    exploded = feedback.explode("theme")
    theme_summary = exploded.groupby("theme", as_index=False).agg(mentions=("comment", "size"), average_rating=("rating", "mean")).sort_values(["mentions", "average_rating"], ascending=[False, True])
    kpis = {"responses": float(len(feedback)), "average_rating": float(feedback["rating"].mean()), "low_rating_rate": float((feedback["rating"] <= 2).mean())}
    return kpis, theme_summary


def main() -> None:
    kpis, themes = analyse_feedback(demo_feedback())
    print({k: round(v, 3) for k, v in kpis.items()})
    print(themes.round(2).to_string(index=False))
    print("Keyword themes are transparent but miss context; review comments manually.")


if __name__ == "__main__":
    main()