UserCF java实现

UserCF是一种基于用户的协同过滤算法，可以用于推荐系统。下面是一个简单的Java实现示例：

1.首先，我们需要定义一个User类来存储用户信息，包括用户ID和评分列表：

public class User {
    private int userId;
    private Map<Integer, Double> ratings;

    public User(int userId) {
        this.userId = userId;
        ratings = new HashMap<>();
    }

    public int getUserId() {
        return userId;
    }

    public void setRating(int itemId, double rating) {
        ratings.put(itemId, rating);
    }

    public Double getRating(int itemId) {
        return ratings.get(itemId);
    }

    public Set<Integer> getRatedItems() {
        return ratings.keySet();
    }
}

2.然后，我们需要定义一个UserCF类来计算用户相似度和进行推荐：

public class UserCF {
    private List<User> users;

    public UserCF(List<User> users) {
        this.users = users;
    }

    // 计算用户相似度
    public double similarity(User u1, User u2) {
        Set<Integer> commonItems = new HashSet<>(u1.getRatedItems());
        commonItems.retainAll(u2.getRatedItems());
        if (commonItems.size() == 0) {
            return 0;
        }

        double sum1 = 0, sum2 = 0, sum3 = 0;
        for (int itemId : commonItems) {
            double r1 = u1.getRating(itemId);
            double r2 = u2.getRating(itemId);
            sum1 += r1 * r2;
            sum2 += r1 * r1;
            sum3 += r2 * r2;
        }

        return sum1 / (Math.sqrt(sum2) * Math.sqrt(sum3));
    }

    // 推荐物品
    public List<Integer> recommend(User user, int n) {
        Map<Integer, Double> scores = new HashMap<>();
        for (User other : users) {
            if (other.getUserId() == user.getUserId()) {
                continue;
            }

            double sim = similarity(user, other);
            if (sim <= 0) {
                continue;
            }

            for (int itemId : other.getRatedItems()) {
                if (user.getRating(itemId) != null) {
                    continue;
                }

                double score = sim * other.getRating(itemId);
                scores.put(itemId, scores.getOrDefault(itemId, 0.0) + score);
            }
        }

        List<Integer> result = new ArrayList<>(scores.keySet());
        result.sort((i1, i2) -> Double.compare(scores.get(i2), scores.get(i1)));

        return result.subList(0, Math.min(n, result.size()));
    }
}

3.最后，我们可以使用UserCF类来进行推荐：

public class Main {
    public static void main(String[] args) {
        List<User> users = new ArrayList<>();
        User user1 = new User(1);
        user1.setRating(1, 4.0);
        user1.setRating(2, 3.0);
        user1.setRating(3, 5.0);
        users.add(user1);

        User user2 = new User(2);
        user2.setRating(2, 4.0);
        user2.setRating(3, 3.0);
        user2.setRating(4, 5.0);
        users.add(user2);

        User user3 = new User(3);
        user3.setRating(1, 5.0);
        user3.setRating(2, 2.0);
        user3.setRating(4, 4.0);
        users.add(user3);

        UserCF userCF = new UserCF(users);
        List<Integer> items = userCF.recommend(user1, 2);
        System.out.println(items);
    }
}

在上面的示例中，我们创建了三个用户，每个用户评分了一些物品。然后，我们使用UserCF类来计算用户相似度并进行推荐。在这个例子中，我们将向用户1推荐两个物品。