Provide a detailed introduction to the latest Facebook algorithm and its weight calculation method.

I will be discussing the algorithm used by Facebook and how weights are calculated for different features. If you wish to repost this content, please provide proper attribution and include a link to this article.

Factors that influence ranking

Affinity Score

The Affinity Score refers to the degree of closeness between a user and content. It is based on the user's interaction history with a post, such as whether they frequently engage with the account or the type of content. The higher the Affinity Score between a user and a post, the higher the post's weight score.

Weight Value

Weight value refers to the user's interaction with a post, including likes, comments, shares, and so on. It is used to measure the weight of a user's interaction with a particular post.

Different types of interactions carry different weights when calculating weight scores. Below are the weight scores for different types of interactions.

Category	Weight score
Like	1
Love	2
Wow	3
Hah	4
Sorry	7
Anger	8
Support	16
Comment	13
Share	26
Popular	400

Weighted Score Algorithm

f1(w1) = P1 x wP1 + P2 x wP2 + P3 x wP3 + P4 x wP4 + P7 x wP7 + C x wC + S x wS

P1-P7 represent the number of likes for different types of likes.
wP1-wP7 represent the weight scores for different types of likes.
C represents the number of comments.
wC represents the weight score for comments.
S represents the number of shares.
wS represents the weight score for shares.

For example, if a post has 19 likes, 8 loves, 3 hugs, 44 comments, and 2 shares, then the weighted score is calculated as 19 * wP1 + 8 * wP2 + 3 * wP3 + 44 * wC + 2 * wS = total score, which in this case is 707. If the total score is greater than 400, then the post is considered to be popular.

Time Decay

Refers to the time of posting a message. As time goes on, the weight score of the message will gradually decrease to reflect its freshness and relevance.

Time Decay Algorithm

decayFactor = 0.5 ^ (age / halfLife)

"Age" represents the time interval between the post's publication and the present, "halfLife" represents the set half-life, which is 12 hours in duration and measured in milliseconds. "DecayFactor" represents the decay factor calculated based on time decay, with a value between 0 and 1.

Homepage Ranking

The total weight score is calculated as the sum of intimacy score and interaction score, multiplied by time decay.

The algorithm will put the posts into the "warehouse" based on the total weight score. Every time a user refreshes the homepage, the posts with higher weight scores will be presented by selecting them from the warehouse.

Post Sentiment

Positive sentiment make posts more likely to be promoted, while negative sentiment decrease the likelihood of post promotion. The following tool can analyze whether the sentiment in a post are positive or negative based on its content.