What are the pitfalls of the recommender system and what is the recommender system

The recommended software is not a personal need, so individuals need to pay attention to whether there is a recommendation in the process of installing the software. This can be canceled, otherwise it will take up space on the computer.

Recommender systems can help users decide what products they should buy, simulating salespeople to help customers through the buying process. Personalized recommendation is to recommend information and products that users are interested in according to their interests and purchase behaviors. With the continuous expansion of e-commerce, the number and variety of products are growing rapidly, and customers need to spend a lot of time to find the products they want to buy.

This process of browsing through a lot of extraneous information and products will undoubtedly lead to a constant loss of consumers who are drowning in the problem of information overload. In order to solve these problems, personalized recommendation systems came into being. The personalized recommendation system is an advanced business intelligence platform built on the basis of massive data mining to help e-commerce** provide fully personalized decision support and information services for its customers' shopping.

The meaning of personalized or rotten recommendation

On the e-commerce platform, personalized recommendations can help users discover their favorite products, thereby improving the shopping experience; On the social platform, personalized recommendations can help users discover the topics and connections they are interested in, so as to expand their social circle. On news platforms, personalized recommendations can help users discover the news content they are interested in, thereby improving the reading experience.

Personalized recommendations require a large amount of user behavior data, so there is a certain risk of privacy leakage. The effectiveness of personalized recommendations is often affected by the recommendation algorithm, and the instability and bias of the algorithm may lead to unsatisfactory recommendation results. In addition, personalized recommendation also requires the platform to have strong algorithm and data processing capabilities, so it may be difficult and costly to achieve personalized recommendation.

A recommender system is all about pushing the "right" content to the "right" people. Specifically, it refers to pushing "information" (including **, **, etc.) to people according to "people's interests" (interests can be used at a point through the person's clicking habits on the page, the duration of page browsing, likes, favorites, and other user behavior data) to people. Transform the idea of people looking for information into the idea of information finding people.

Recommendations are the result of data mining based on the user's behavioral data. Without data, no valuable mining will be generated. Data includes: enough users, enough behavioral data, and enough content.

As shown in the figure below, **A and **B are both liked by user B, user C and user D, so it is judged that **A and **B are similar, and if user A likes **A, **B similar to **A will be recommended to User A.

After recall-coarse-fine-rearranged, finally for each user, a list of final estimated scores has been laid [content, score], when the user starts the app, the client sends a request to the server, and the server calls the recommendation algorithm interface to recommend the content to the user in descending order of the estimated score.