Tiktok algorithm and block chain

1.

1、 Recommendation principle of intelligent algorithm

the essence of intelligent algorithm recommendation is to match the most interesting content for current users from an aggregate content pool

this content pool contains tens of millions of content every day, covering 15s short video, 1min long video and 5min super long video

when matching content to users, the platform mainly depends on three factors: content, users and their interest in content

How does the

system understand our creative content

when describing the content, the platform will mainly rely on keyword recognition technology: by extracting the keywords in writing and video, the content will be roughly classified according to the keywords, and then the classification will be refined according to the keywords in the subdivision field

for example, the keywords of video and content are "Ronaldo, football, World Cup"

most of the key words belong to sports vocabulary. You will first classify your works into sports categories, and then subdivide them into "football", "international football" and other secondary and tertiary categories according to the specific key words

user characterization

through this series of comparison and analysis, the system speculates and restores the basic attributes of a user, for example: TA may be a male who is traveling and likes football, car and other categories

the system will classify the above user characteristics as the user's tag

User tags are mainly divided into three categories:

1) basic information of users (age, gender, region)

2) user's behavior information (attention account number, history wandering record, like collection content, music, topic)

3) reading interest (reading behavior, user clustering, user tagging)

according to the user's information and behavior, the system analyzes and calculates the user's classification, topics, people and other information, so as to complete the system's characterization of users

the essence of recommendation algorithm

uses the characteristics of works (keywords, labels, popularity, forwarding, timeliness, similarity), user preferences (short-term click behavior, interest, occupation, age, gender, etc.), and environmental factors (region, time, weather, network environment) to fit a function of user satisfaction with content, It will estimate the user's click probability for each work, and then rank all the works according to their interests from tens of millions of content flow pools in the system. The top 10 works will stand out at this time and be recommended to the user's mobile phone for display

it's probably like this. If you want to learn, you can write to Xiaobian in private

2. Tiktok algorithm is actually a funnel mechanism, which is basically consistent with the principle of recommendation algorithm of the headline's de centralization. Tiktok tiktok is divided into three steps:

first, cold start flow cell exposure

suppose that 1 million people upload short videos every day on the shaking, and the jitter will randomly assign each cold video to a cold start flow pool with an average exposure. For example, after each short video is sent out by auditing, there are 1000 exposures

second on average. Data selection

tiktok will analyze 1000 points of the 1 million short videos, analyze the data of each dimension such as point, attention, comment, forwarding and so on, and then pick out videos with more than 10% indexes, and then distribute 100 thousand times on average. Then go to see which are like, follow, forward and comment more than 10%, and roll into the next round of larger flow pool for recommendation

thirdly, the boutique recommendation pool

through round after round of verification, short videos with extremely high liking rate, playback completion rate, comment interaction rate and other indicators are screened out to have the opportunity to enter the boutique recommendation pool. When users open the pool, they see videos with tens of millions of likes< br />
: " We are paladins. We can't let revenge occupy our consciousness& quot;,

3. Now it's just testing, there's no official release, there's no use for it?

4.

1. Machine audit + manual double audit

when a video is initially uploaded, the platform will give you an initial flow. If after the initial flow, according to the like rate, comment rate and forwarding rate, judge whether the video is popular or not. If the first round of evaluation is popular, then it will carry out the second transmission

when you get the optimal feedback for the second time, you will get more traffic

on the contrary, in the first wave or the nth wave, if the response is not good, it will no longer be recommended. Without the recommendation of the platform, the probability of your video wanting to fire is very small, because there is no more traffic to see you. The first step of video fire is to be seen by others, the first step is to die, and the follow-up can only rely on friends' praise

the logic behind this algorithm: intelligent distribution, overlay recommendation, and heat weighting

2, overlay recommendation

the so-called overlay recommendation refers to the new video will intelligently distribute about 100vv of playback, if the forwarding amount reaches 10 (for example), the algorithm will judge as popular content, automatically weight the content, and overlay recommendation will give you 1000vv; The forwarding amount is 100 (for example), and the algorithm continues to stack up to 10000 vv; The tiktok is 1000 (for example), and then it is recommended to 10wvv, and what's more, the number of players who play the night is also more than the number of players. p>

overlay recommendation is of course based on the comprehensive weight of the content as the evaluation standard. The key indicators of the comprehensive weight are: the completion rate, the amount of likes, the amount of comments, and the amount of forwarding, and the weight of each step is different. When it reaches a certain level, the mechanism of combining big data algorithm and manual operation is adopted

3, tiktok

tiktok nearly 100 exploding flare, found all the night exploding video, and the video of the recommended voice board, the volume of the play is more than one million, and the comprehensive data (broadcast rate, point praise, comment quantity, forwarding volume) are all good except. p>

extended data:

1: improve your own data, the more complete the better. Including avatars, nicknames, mobile phones, microblogs, wechat, headlines, etc., the more detailed the better. Because it is machine and manual double audit, once the machine audit, will carry out a large number of inferior elimination

2. The video should be bright. The video is only 15 seconds. In this short 15 seconds, there is no highlight and no turning point. You will not have any interaction with you, and there is also a shielding function. Once the user blocks you, this is a very serious thing, because the user will not be recommended for your short video in the future

5. Talk about love

6. The content of blockchain science popularization can be published, and other evidences are not clear. The application fields of blockchain include digital currency, token, finance, anti-counterfeiting traceability, privacy protection, supply chain, entertainment, etc. with the popularity of blockchain and bitcoin, many related top domain names have been registered, which has a great impact on the domain name instry.

7. Tiktok is short video APP, but it's different from fast video and jitter, and Kwai can earn money! Tiktok is just killing time. Playing Kwai ask video can really make money. Pro test has been made up!

8. After uploading the video, the machine recommends it to the people who may be interested in your video tags in a small range, that is, a small flow pool. It's about 20-250 people (including contacts friends, fans of accounts, fans of music accounts, fans who like or pay attention to this topic, fans in the same city and randomly allocated by the system)
calculate the number of comments, reviews and shares of the audience in unit time
the specific formula is: popularity = a number of comments + B number of likes + C number of shares, coefficient a, B, C will fine tune in real time according to the overall algorithm, roughly: C & gt; A> B This step is called the first recommendation. This is why we usually see the content in the recommendation, some of the interaction rate is almost 0. Because you're the first audience of this video.