python - 有没有轻量级的推荐系统？

Question

基础数据量在十万级别，每天大概有三五百条的增量。

有没有针对这种小型数据量的推荐系统？或者这种根本就不需要推荐系统？

但是也需要个性化推送。

或者有什么别的方法？

Answer 1

As far as I know and looking at the python official website, there is no one as mature as request and scrapy, or not at all.

I recommend (I have read it) that you can read the recommendation system chapter of the book "Collective Programming Wisdom" (pages 20-30) to implement a simple recommendation system
(The book is practical, almost There is no need to learn relevant theoretical knowledge. One example in the book is to recommend movies to people based on movie ratings)

The level of 100,000 is a very small order of magnitude, a simple algorithm is enough, and there is no need to consider performance.

Sorry, I looked carefully and found that there are still some, but I don’t know if they are useful or not

https://www.tensorflow.org/
http://caffe.berkeleyvision.org/

Answer 2

You can refer to the following articles (need to circumvent the firewall), which mainly include Crab, pysuggest 1.0, unison-recsys and python-recsys.
Python libraries for building recommender systems

Answer 3

Actually, I think it’s better to write one for this scenario or host it on Amazon cloud host. It’s also very cheap. If you write it yourself, you can also practice the application of machine learning algorithms