Samamon Khemmarat

The impact of YouTube recommendation system on video views

Hosting a collection of millions of videos, YouTube offers several features to help users discover the videos of their interest. For example, YouTube provides video search, related video recommendation and front page highlight. The understanding of how these features drive video views is useful for creating a strategy to drive video popularity. In this paper, we perform a measurement study on data sets crawled from YouTube and find that the related video recommendation, which recommends the videos that are related to the video a user is watching, is one of the most important view sources of videos. Despite the fact that the YouTube video search is the number one source of views in aggregation, the related video recommendation is the main source of views for the majority of the videos on YouTube. Furthermore, our results reveal that there is a strong correlation between the view count of a video and the average view count of its top referrer videos. This implies that a video has a higher chance to become popular when it is placed on the related video recommendation lists of popular videos. We also find that the click through rate from a video to its related videos is high and the position of a video in a related video list plays a critical role in the click through rate. Finally, our evaluation of the impact of the related video recommendation system on the diversity of video views indicates that the current recommendation system helps to increase the diversity of video views in aggregation.

Boosting video popularity through recommendation systems

While search engines are the major sources of content discovery on online content providers and e-commerce sites, their capability is limited since textual descriptions cannot fully describe the semantic of content such as videos. Recommendation systems are now widely used in online content providers and e-commerce sites and play an important role in discovering content. In this paper, we describe how one can boost the popularity of a video through the recommendation system in YouTube. We present a model that captures the view propagation between videos through the recommendation linkage and quantifies the influence that a video has on the popularity of another video. Furthermore, we identify that the similarity in titles and tags is an important factor in forming the recommendation linkage between videos. This suggests that one can manipulate the metadata of a video to boost its popularity.

How YouTube videos are discovered and its impact on video views

As the largest video sharing site around the world, YouTube has been changing the way people entertain, gain popularity, and advertise. Discovering the major sources that drive views to a video and understanding how they impact the view growth pattern have become interesting topics for researchers as well as advertisers, media companies, or anyone who wish to have a shortcut to stardom. The work of this paper is to identify three major view sources, related video recommendation, YouTube search, and video highlight such as popular video list on YouTube homepage or video embedding on social networking sites, and examine the patterns of views from each view source. First, the impact of each view source on the view diversity and on the view share of each individual video is analyzed. It is found that while search and highlight create an effect of rich-get-richer, the related video recommendation equalizes the view distribution and helps users find niche videos. Second, the contribution of the three view sources to video popularity growth is investigated. The investigation reveals that search and related video recommendation are the two major sources that persistently drive views to a video. The view rates from recommendation and search are generally stabilized to be constant view rates. Third, the underlying factors that affect the long-term view rate from referrer videos are explored. The results indicate that the top referrer video set of a video is fairly stable and the view rate from recommendation is mainly determined by view rates of top referrer videos. Finally, whether highlight increases the view rate of a video after the duration of promotion is studied. The observations suggest that video highlight does not directly impact the view rate of a video after the event finishes. The findings presented in the paper provide several key insights into the impact and patterns of view contributions for each major source of the video views.

Planet YouTube: Global, measurement-based performance analysis of viewer;'s experience watching user generated videos

User experience is a very important aspect of user-generated video streaming service such as YouTube. In this paper, we perform a global study of user experience for YouTube videos using PlanetLab nodes from all over the world. We analyze the number of pauses, accumulative pause time, rate at which videos were downloaded to clients, and how the YouTube infrastructure impact the viewers experience. The data for this analysis was generated by an automated tool from the traces captured during the PlanetLab-based measurement. Results from this analysis show that on average there are about 2.5 pauses per video. Interestingly, we found that on average 25% of the videos with pauses have an accumulative pause time greater than 15 seconds. This shows that not only the number of pauses but also the total length of pauses has to be analyzed to investigate the user experience of watching videos offered by YouTube. In addition, our results show that there is an inverse relationship between the download rates of the videos and the number of pauses encountered.

Fast top-k path-based relevance query on massive graphs

The task of obtaining the items highly-relevant to a given set of query items is a basis for various applications, such as recommendation and prediction. A family of path-based relevance metrics, which quantify item relevance based on the paths in a given item graph, have been shown to be effective in capturing the relevance in many applications. Despite their effectiveness, path-based relevance normally requires time-consuming iterative computation. We propose an approach to obtain the top-k most relevant items for a given query item set quickly. Our approach can obtain the top-k items without having to compute converged scores. The approach is designed for a distributed environment, which makes it scale for massive graphs having hundreds of millions of nodes. Our experimental results show that the proposed approach can produce the result 20 to 50 times faster than a previously proposed approach and can scale well with both the size of input and the number of machines used in the computation.

Fast Top-K Path-Based Relevance Query on Massive Graphs

Obtaining the items highly-relevant to a given set of query items is a key task for various applications, such as recommendation and relationship prediction. A family of path-based relevance metrics, which quantify item relevance based on the paths in an item graph, have been shown to be effective in capturing the relevance in many applications. Despite their effectiveness, path-based relevance normally requires time-consuming iterative computation. We propose an approach to obtain the top-k most relevant items for a given query item set quickly. Our approach uses novel score bounds to detect the emergence of the top-k items during the computation. The approach is designed for a distributed environment, which makes it scale for massive graphs having billions of nodes. Our experimental results show that the proposed approach can provide the results up to two order of magnitudes faster than previously proposed approaches and can scale well with both the size of input and the number of machines used in the computation.

Querying Web-Scale Information Networks Through Bounding Matching Scores

Web-scale information networks containing billions of entities are common nowadays. Querying these networks can be modeled as a subgraph matching problem. Since information networks are incomplete and noisy in nature, it is important to discover answers that match exactly as well as answers that are similar to queries. Existing graph matching algorithms usually use graph indices to improve the efficiency of query processing. For web-scale information networks, it may not be feasible to build the graph indices due to the amount of work and the memory/storage required. In this paper, we propose an efficient algorithm for finding the best k answers for a given query without precomputing graph indices. The quality of an answer is measured by a matching score that is computed online. To speed up query processing, we propose a novel technique for bounding the matching scores during the computation. By using bounds, we can efficiently prune the answers that have low qualities without having to evaluate all possible answers. The bounding technique can be implemented in a distributed environment, allowing our approach to efficiently answer the queries on web-scale information networks. We demonstrate the effectiveness and the efficiency of our approach through a series of experiments on real-world information networks. The result shows that our bounding technique can reduce the running time up to two orders of magnitude comparing to an approach that does not use bounds.

On Understanding User Interests through Heterogeneous Data Sources

User interests can be learned from multiple sources, each of them presenting only partial facets. We propose an approach to merge user information from disparate data sources to enable a more complete, enriched view of user interests. Using our approach, we show that merging different sources results in three times of more interest categories in user profiles than with each single source and that merged profiles can capture much more common interests among a group of users, which is key to group profiling.

A distributed approach for top-k star queries on massive information networks

Massive information networks, such as the knowledge graph by Google, contain billions of labeled entities. Star queries, which aim to identify an entity, given a set of related entities, are common on such networks. Answering star queries can be modeled as a graph pattern matching problem. Traditional approaches apply graph indices to accelerate the query processing. Unfortunately, it is so costly that it is nearly infeasible to build indices on billion node graphs since the time or storage complexity of most indexing techniques is super-linear to the graph size. In this paper, we propose an algorithm to identify the top-k best answers for a star query. Instead of using expensive indices, our algorithm utilizes novel bounding techniques to derive the top-k best answers efficiently. Further, the algorithm can be implemented in a distributed manner scaling to billions of entities and hundreds of machines. We demonstrate the effectiveness and the efficiency of our approach through a series of experiments on real-world information networks.

Boosting video popularity through keyword suggestion and recommendation systems

YouTube offers a great opportunity for people to entertain, advertise, gain popularity, and generate revenue. How to increase views for a video has become the key question for anyone who wish to be famous or gain more revenue. Recognizing that a recommendation system is a major view source for videos, our goal in this paper is to increase video views in YouTube by leveraging on the recommendation system. We first perform measurements to understand how views are propagated among videos through recommendation links and identify factors that influence the recommendation produced by the system. Our measurement results show that similarity in video meta-data is a crucial ingredient in connecting videos. We then propose a keyword suggestion method for a video with the aim to raise video views through the recommendation system. The keyword suggestion method utilizes video clusters on a referrer video graph to obtain relevant keywords and ranks keywords based on both their relevance and their potential to attract video views. The effectiveness of the keyword suggestion method is demonstrated through a case study, showing that using the keywords suggested by our method leads to a larger number of video views and higher average watching time per video playback compared to initial user-given keywords.