根据用户对 Django Rest Framework 的兴趣，有没有可靠的方法从 PostgreSQL 获取数据？答案

【问题标题】：Is there any reliable way to get the data from PostgreSQL according to the user interest in the Django Rest Framework?根据用户对 Django Rest Framework 的兴趣，有没有可靠的方法从 PostgreSQL 获取数据？
【发布时间】：2022-11-03 18:36:20
【问题描述】：

我们的项目在功能上类似于 Twitter 我们希望根据用户对帖子的点赞、分享和cmets，根据用户的兴趣在Feed上显示帖子

我们可以如何以及在哪里保存用户的活动？
我们如何使用 Django QuerySet 和 DRF 为用户分页过滤和获取帖子数据
我们如何使用最近+兴趣相关的数据对帖子数据进行排序
如果有任何更好的方法存在，那么请分享一些关于它的细节。

提前致谢！

【问题讨论】：

在我看来，最好的方法是为此目的编写自己的算法。

标签： django postgresql django-rest-framework

【解决方案1】：

This is a standard personalization/recommendation use case, and there are a few standard approaches:

Content-based filtering: select posts that are similar to ones the user has interacted with. You can use techniques like TF-IDF to calculate a similarity score between posts.
Collaborative filtering: select posts that people with a similar activity history to the user have interacted with.

Typically you would precalculate recommendation scores in a background job, and store them in a separate table. Then when showing a feed to a user, you just ORDER BY the score column to prioritize recommended posts.
You can combine the two approaches by taking a weighted average of content-based and collaborative scores.
Note that all of this assumes you have a adequate volume of data and activity to make the recommendations useful. For a new site with few users it may be better to just show chronological posts initially.

【讨论】：