Function update_profile_based_on_selection

agents/feedback-agent/main.py:542–585

Update the profile via the unified drift-aware updater.

(
    user_id: str,
    selected_paper_ids: List[int],
    papers: List[Dict],
    skipped_paper_ids: Optional[List[int]] = None,
    historical_selected_papers: Optional[List[Dict[str, Any]]] = None,
    current_timestamp: Optional[datetime] = None,
    feedback_strength_multiplier: float = 1.0,
)

def update_profile_based_on_selection(
    user_id: str,
    selected_paper_ids: List[int],
    papers: List[Dict],
    skipped_paper_ids: Optional[List[int]] = None,
    historical_selected_papers: Optional[List[Dict[str, Any]]] = None,
    current_timestamp: Optional[datetime] = None,
    feedback_strength_multiplier: float = 1.0,
) -> Dict:
    """Update the profile via the unified drift-aware updater."""
    profile = get_profile(user_id)
    if not profile:
        print(f"Warning: No profile found for user {user_id}")
        return {}

    now = current_timestamp or datetime.now()
    normalized_profile = ensure_profile_schema(profile, now=now)

    selected_papers: List[Dict[str, Any]] = []
    for paper_number in selected_paper_ids:
        idx = paper_number - 1
        if 0 <= idx < len(papers):
            selected_paper = dict(papers[idx])
            selected_paper["authors"] = normalize_authors(selected_paper.get("authors", []))
            selected_papers.append(selected_paper)

    skipped_papers: List[Dict[str, Any]] = []
    for paper_number in skipped_paper_ids or []:
        idx = paper_number - 1
        if 0 <= idx < len(papers):
            skipped_paper = dict(papers[idx])
            skipped_paper["authors"] = normalize_authors(skipped_paper.get("authors", []))
            skipped_papers.append(skipped_paper)

    updated_profile = update_profile_with_feedback(
        normalized_profile,
        selected_papers,
        skipped_papers,
        historical_selected_papers=historical_selected_papers or [],
        current_time=now,
        feedback_strength_multiplier=feedback_strength_multiplier,
    )
    update_profile(user_id, updated_profile)
    return updated_profile


def estimate_feedback_strength_multiplier(push_id: str, current_timestamp: datetime) -> tuple[float, Optional[float]]: