'''
Write a recommender function in Python, the input arguments are a list of string represents user preferred keywords and a list of articles with their own keywords, the output is the articles sorted by the relevance to the user preferred keywords.
'''
def recommend(user_keywords, articles):
    # Calculate relevance scores for each article based on user's preferred keywords
    scores = []
    for article in articles:
        common_keywords = set(article["keywords"]).intersection(user_keywords)
        score = len(common_keywords) / len(user_keywords)
        scores.append((article, score))

    # Sort articles by relevance score
    sorted_articles = sorted(scores, key=lambda x: x[1], reverse=True)

    # Return list of articles in descending order of relevance
    return [article[0] for article in sorted_articles]