Discovering and accessing specific content within educational video bases is a challenging task, mainly because of the abundance of video content and its diversity. Recommender systems are often used to enhance the ability to find and select content. But, recommendation mechanisms, especially those based on textual information, exhibit some limitations, such as being error-prone to manually created keywords or due to imprecise speech recognition. This paper presents a method for generating educational video recommendation using deep face-features of lecturers without identifying them. More precisely, we use an unsupervised face clustering mechanism to create relations among the videos based on the lecturer's presence. Then, for a selected educational video taken as a reference, we recommend the ones where the presence of the same lecturers is detected. Moreover, we rank these recommended videos based on the amount of time the referenced lecturers were present. For this task, we achieved a mAP value of 99.165%.

This paper proposes a generative method to dynamically simulate users' short video watching journey for watch time prediction in short video recommendation. Unlike existing methods that rely on multimodal features for video content understanding, our method simulates users' sustained interest in watching short videos by learning collaborative information, using interest changes from existing positive and negative feedback videos and user interaction behaviors to implicitly model users' video watching journey. By segmenting videos based on duration and adopting a Transformer-like architecture, our method can capture sequential dependencies between segments while mitigating duration bias. Extensive experiments on industrial-scale and public datasets demonstrate that our method achieves state-of-the-art performance on watch time prediction tasks. The method has been deployed on Kuaishou Lite, achieving a significant improvement of +0.13\% in APP duration, and reaching an XAUC of 83\% for single video watch time prediction on industrial-scale streaming training sets, far exceeding other methods. The proposed method provides a scalable and effective solution for video recommendation through segment-level modeling and user engagement feedback.

Watch time prediction (WTP) has emerged as a pivotal task in short video recommendation systems, designed to encapsulate user interests. Predicting users' watch times on videos often encounters challenges, including wide value ranges and imbalanced data distributions, which can lead to significant bias when directly regressing watch time. Recent studies have tried to tackle these issues by converting the continuous watch time estimation into an ordinal classification task. While these methods are somewhat effective, they exhibit notable limitations. Inspired by language modeling, we propose a novel Generative Regression (GR) paradigm for WTP based on sequence generation. This approach employs structural discretization to enable the lossless reconstruction of original values while maintaining prediction fidelity. By formulating the prediction problem as a numerical-to-sequence mapping, and with meticulously designed vocabulary and label encodings, each watch time is transformed into a sequence of tokens. To expedite model training, we introduce the curriculum learning with an embedding mixup strategy which can mitigate training-and-inference inconsistency associated with teacher forcing. We evaluate our method against state-of-the-art approaches on four public datasets and one industrial dataset. We also perform online A/B testing on Kuaishou, a leading video app with about 400 million DAUs, to demonstrate the real-world efficacy of our method. The results conclusively show that GR outperforms existing techniques significantly. Furthermore, we successfully apply GR to another regression task in recommendation systems, i.e., Lifetime Value (LTV) prediction, which highlights its potential as a novel and effective solution to general regression challenges.

Within the domain of short video recommendation, predicting users' watch time is a critical but challenging task. Prevailing deterministic solutions obtain accurate debiased statistical models, yet they neglect the intrinsic uncertainty inherent in user environments. In our observation, we found that this uncertainty could potentially limit these methods' accuracy in watch-time prediction on our online platform, despite that we have employed numerous features and complex network architectures. Consequently, we believe that a better solution is to model the conditional distribution of this uncertain watch time. In this paper, we introduce a novel estimation technique -- Conditional Quantile Estimation (CQE), which utilizes quantile regression to capture the nuanced distribution of watch time. The learned distribution accounts for the stochastic nature of users, thereby it provides a more accurate and robust estimation. In addition, we also design several strategies to enhance the quantile prediction including conditional expectation, conservative estimation, and dynamic quantile combination. We verify the effectiveness of our method through extensive offline evaluations using public datasets as well as deployment in a real-world video application with over 300 million daily active users.

YouTube relies heavily on AI to deliver content. The newest YouTube algorithms put a great deal of value on the average time that a person views any video, gives it a like or dislike, and comments. Similarly, the recommender system is one of the most powerful use cases of ML that is encountered by every one of us many times a day. Collaborative Filtering: This is a type where we tend to build collaborations between various users and items(videos). Matrix Factorization: It tries to dissolve both user and item vectors together thus decomposing them and providing YouTube with better comparison metrics.

Be responsible for the entire algorithmic life-cycle in the company: data analytics, prototyping of new ideas, implementing algorithms in a production environment and then monitoring and maintaining them Turn algorithm prototypes into shippable products that will have a significant and immediate impact on the company's revenue Work on a daily basis with some of the hottest trends in today's job market: machine/deep learning, big data analytics and cloud computing Apply your scientific knowledge and creativity to analyze large volumes of diverse data and develop algorithms to solve complex problems Influence directly on the way billions of people discover the internet Work on projects such as Internet Personalization, Content Feed, Real Time Bidding, Video Recommendations and much more Turn algorithm prototypes into shippable products that will have a significant and immediate impact on the company's revenue

Many enterprises have been working with Apache Spark and ML algorithms for improved results. Yahoo, for example, uses Apache Spark along with ML algorithms to collect innovative topics than can enhance user interest. If only ML is used for this purpose, over 20, 000 lines of code in C or C will be needed, but with Apache Spark, the programming code is snipped at 150 lines! Another example is Netflix where Apache Spark is used for real-time streaming, providing better video recommendations to users. Streaming technology is dependent on event data, and Apache Spark ML facilities greatly improve the efficiency of video recommendations.

In a rare public feud, Amazon has said that its Echo Show devices could no longer play videos from YouTube because the site's parent, Google, stopped supporting the service. While Amazon claims that there is'no technical reason' for the decision, Google says that Amazon's use of YouTube on the Echo Show created a'broken user experience.' Experts are describing the removal of YouTube as a'bit of a blow' to Amazon, as a big chunk of the possible video content you could watch on Echo Show is now gone. In a rare public feud, Amazon has said that its Echo Show devices could no longer play videos from YouTube because the site's parent, Google, stopped supporting the service While Amazon claims that there is'no technical reason' for the decision, Google said this wasn't the case. 'We've been in negotiations with Amazon for a long time, working towards an agreement that provides great experiences for customers on both platforms,' it said in a statement.

Video publishers have been caught off guard with the recent announcement of Apple blocking video autoplay. Even Google is pushing back on bad web ads. The backlash against video autoplay has been festering for some time. If losing video ad revenue and turning consumers off with declining traffic isn't a wake-up call then what will be? Headlines like this from CNN "Apple's plan to kill autoplay feature could leave publishers in the dust" should get video publisher's attention.

You'll soon be getting much better video recommendations from YouTube. Google is launching new iOS and Android apps for YouTube aimed not only at a design refresh, but also at offering you smarter video recommendations. You'll also notice larger photos, which should help in identifying videos from particular channels. "The new recommendation system is based on deep neural network technology, which means it can find patterns automatically and keep learning and improving as it goes," said a blog post detailing the update. "Every day, we recommend hundreds of millions of different videos on Home, billions of times, in 76 languages."