TikTok recently published in their press release article what factors make up the algorithm in the for you page. They explain that certain criteria are more important than others and they are all valued differently. They mention that criteria include:
TikTok近期在他們新聞稿集里,發(fā)布了關(guān)于For You頁面的算法構(gòu)成要素。他們解釋到,某些標(biāo)準(zhǔn)比其他的更重要,并且(相同的)標(biāo)準(zhǔn)會(huì)被估算成不同的權(quán)重。
他們提到的標(biāo)準(zhǔn)包括:
The user’s engagement with the post. Whether this is likes, comments, shares or following the account; these are all given a different importance
用戶對(duì)發(fā)布(行為或內(nèi)容,也可以簡(jiǎn)單的理解成帖子)的參與程度。不論是點(diǎn)贊,評(píng)論,分享或者關(guān)注賬戶;它們都會(huì)被給予不同的重要性。
Which sounds are used. This answers the predicted theory that trending sounds boosts your video’s reach, hence demonstrating the importance of using trending sounds
使用哪些音效。這個(gè)回應(yīng)了一個(gè)已被預(yù)測(cè)的理論,(即)流行的音效增加了你的視頻的搜索,因此,闡述了使用流行音效的重要性。
The device and location settings. The language your device is using, GPS location and type of device are also considered within the algorithm
設(shè)備和位置設(shè)置。設(shè)備所使用的語言、GPS位置和設(shè)備類型也會(huì)在算法中被考慮。
It was also mentioned that each one of these indicators is valued differently for each different user, this is due to different users having different habits, for example, if a user only shares one out of every 200 videos, the algorithm is going to value that share a lot more than people who share every second video.
它還提到了,這些指標(biāo)中的每一個(gè)指標(biāo)都會(huì)因?yàn)椴挥脩舯旧淼氖褂昧?xí)慣不盡相同,而產(chǎn)生不同權(quán)重;例如,如果某用戶只在兩百個(gè)視頻中分享了一個(gè)(出去),那么算法就會(huì)認(rèn)為這個(gè)分享(的權(quán)重)比其他什么都分享的用戶(的權(quán)重)高很多。
Something very important to mention is how a high number of followers “is not a direct factor in the algorithm”. This means that even if you have a very small number of followers, if it performs well against the criteria mentioned above, it will be treated the same as if it was done by an account with millions of followers.
有一點(diǎn)非常重要,巨大的粉絲數(shù)量不是平臺(tái)推薦算法機(jī)制的直接加權(quán)項(xiàng)。這意味著,即使你賬戶的粉絲數(shù)量非常少,如果按照上述特定標(biāo)準(zhǔn)中表現(xiàn)良好,將會(huì)被視為與同樣標(biāo)準(zhǔn)同樣表現(xiàn)的數(shù)百萬粉絲賬戶的重要性一致。
The article also mentions how there are numerous factors that are involved to understand what a user is NOT interested in. This includes:
文章還提到,大量有關(guān)什么是用戶不感興趣的要素是怎么關(guān)聯(lián)的。這些因素包括
- 舉報(bào)視頻
- Reporting the video
- 點(diǎn)擊 "我不感興趣的按鈕"
- Tapping the “I’m not interested button”
- 取消關(guān)注該用戶
- Unfollowing the user
It is clearly mentioned that you will not be recommended content that you have already seen before or any duplicate content, as TikTok considers this spam. This highlights the importance of creating original content, as copying and reposting someone else’s video will drastically decrease the video’s performance.
其中明確提到,不會(huì)給你推薦之前已經(jīng)看過的內(nèi)容,也不會(huì)推薦任何重復(fù)的內(nèi)容,因?yàn)門ikTok認(rèn)為這是垃圾。這就凸顯了創(chuàng)作原創(chuàng)內(nèi)容的重要性,因?yàn)榘徇\(yùn)和轉(zhuǎn)發(fā)別人的視頻會(huì)徹底地降低視頻的表現(xiàn)分。
TikTok finally fully explained the reason behind why you often find videos on the for you page with less than 100 likes. This is due to “bringing diversity into your feed” to “give you opportunities to find new content categories and creators”. This is why the number of followers on someone’s account does not have any relationship to how well the video performs against the algorithm.
TikTok終于充分解釋了,為什么你經(jīng)常在For You頁面上發(fā)現(xiàn)的視頻,只有不到100個(gè)贊背后的原因。這是由于 "要保持你的Feed的多樣性",以便 "讓你有機(jī)會(huì)找到新的內(nèi)容類性和創(chuàng)作者"。這就是為什么用戶賬戶的粉絲數(shù)量,不會(huì)影響算法對(duì)該視頻好壞的判斷。
Lastly, they mention that the for you algorithm “is designed to continuously improve, correct and learn”. Therefore, it must be considered that whilst the factors mentioned above are important in determining the reach of a video, this might change in the future.
最后,它提到For You算法 "旨在不斷改進(jìn)、糾正和學(xué)習(xí)"。因此,必須考慮到,雖然上述因素在決定視頻的播放范圍方面很重要,但是在未來仍然會(huì)產(chǎn)生變化。
原文從8個(gè)方面來闡述TikTok中For You的推送機(jī)制。
1,F(xiàn)or You條件因素
一個(gè)很重的標(biāo)準(zhǔn),就是用戶的“喜好”。
1)用戶互動(dòng):喜好和分享的視頻,關(guān)注的賬戶,發(fā)布的評(píng)論以及創(chuàng)建的內(nèi)容。
2)視頻信息,視頻的標(biāo)題,配音以及標(biāo)簽等細(xì)節(jié)。
3)設(shè)備和賬戶的設(shè)置,語言的偏好,國(guó)家的設(shè)置和設(shè)備的設(shè)置(偽裝流的核心關(guān)注點(diǎn):原文引用:Device and account settings like your language preference, country setting, and device type. These factors are included to make sure the system is optimized for performance.)。
這些因素,都會(huì)被TikTok的推薦系統(tǒng)處理,并依據(jù)這些對(duì)用戶的價(jià)值進(jìn)行加權(quán)。一個(gè)強(qiáng)的興趣指標(biāo),會(huì)比一個(gè)弱的興趣指標(biāo)獲得更大的權(quán)重,比如說,用戶是否會(huì)從頭到尾,完整的看完一個(gè)視頻。是否會(huì)完整的看完一個(gè)長(zhǎng)視頻。
2,F(xiàn)or You Feed的建立
一個(gè)新的用戶,創(chuàng)建賬號(hào)同時(shí),平臺(tái)會(huì)為新用戶,推薦感興趣的類別,以依據(jù)這些喜好,開始進(jìn)行推薦。并依據(jù)此類喜好,進(jìn)行客戶推薦的打磨。
那么對(duì)于沒有選擇喜好的用戶,平臺(tái)會(huì)先推薦熱門視頻,新用戶的第一次點(diǎn)贊,評(píng)論和重播都將是平臺(tái)更加的了解新的用戶,并開始推薦機(jī)制。
每一次用戶的新的互動(dòng)都是讓系統(tǒng)更了解用戶的需求,并推薦更多用戶喜好的視頻。那么當(dāng)用戶選擇“Discover” 選項(xiàng)卡上的標(biāo)簽、聲音、效果和熱門話題也是如此,這些內(nèi)容,都會(huì)進(jìn)入用戶的新的推薦機(jī)制選擇項(xiàng)中。
3,少看自己不感興趣的東西
用戶遇到不感興趣的話題時(shí),可選擇“不感興趣”來選擇。也可選擇隱藏來自特定創(chuàng)作者或者某種背景音樂制作的視頻等方式,來減少系統(tǒng)中的推薦。
4,來自推薦引擎的挑戰(zhàn)
給出“Filter Bubble”(字母翻譯過濾泡沫)的概念。提出挑戰(zhàn):可能會(huì)有更多同質(zhì)化的內(nèi)容被推薦出來。
5,打斷重復(fù)模式
將不同類型的內(nèi)容與用戶喜愛的內(nèi)容穿插在一起,可能會(huì)被推薦一個(gè)受到其他興趣相投的用戶好評(píng)的視頻。但,不推薦重復(fù)的內(nèi)容,你已經(jīng)看過的內(nèi)容,或者任何被認(rèn)為是垃圾的內(nèi)容。
6,多元化內(nèi)容提供
會(huì)在用戶的信息源中提供該用戶的興趣無關(guān)的視頻,或者是積累大量的贊的優(yōu)質(zhì)視頻。是為了讓用戶有更多的機(jī)會(huì)在瀏覽Feed時(shí),發(fā)現(xiàn)新的內(nèi)容類別,發(fā)現(xiàn)新的創(chuàng)作者,并體驗(yàn)新的觀點(diǎn)和想法。(此點(diǎn)也是官方不斷闡述原創(chuàng),優(yōu)質(zhì)視頻的重要性。)
7,保障用戶體驗(yàn)
拒絕敏感視頻,拒絕剛剛上傳或正在審核的視頻,以及尋求人為增加流量的視頻等垃圾內(nèi)容。
8,改進(jìn)內(nèi)容
系統(tǒng)的設(shè)計(jì)是為了不斷改進(jìn)、糾正,根據(jù)用戶的反饋、研究和數(shù)據(jù),努力完善準(zhǔn)確性、調(diào)整模型,并重新評(píng)估有助于推薦的因素和權(quán)重。
我們走過了1年零4個(gè)月的探尋和摸索,目前,我們擁有來我們的跨境中國(guó)團(tuán)隊(duì),北美團(tuán)隊(duì),澳洲團(tuán)隊(duì)。關(guān)于TikTok短視頻流量變現(xiàn)問題及情況。我們會(huì)實(shí)時(shí)分享我們的工作經(jīng)驗(yàn) 。
有對(duì)tiktok 有興趣,有研究的小伙伴,歡迎留言討論。(來源:編譯巨浪出海David)
以上內(nèi)容屬作者個(gè)人觀點(diǎn),不代表雨果網(wǎng)立場(chǎng)!本文經(jīng)原作者授權(quán)轉(zhuǎn)載,轉(zhuǎn)載需經(jīng)原作者授權(quán)同意。