In this study, we present a meta-learning model to adapt the predictions of the network's capacity between viewers who participate in a live video streaming event. We propose the MELANIE model, where an event is formulated as a Markov Decision Process, performing meta-learning on reinforcement learning tasks. By considering a new event as a task, we design an actor-critic learning scheme to compute the optimal policy on estimating the viewers' high-bandwidth connections. To ensure fast adaptation to new connections or changes among viewers during an event, we implement a prioritized replay memory buffer based on the Kullback-Leibler divergence of the reward/throughput of the viewers' connections. Moreover, we adopt a model-agnostic meta-learning framework to generate a global model from past events. As viewers scarcely participate in several events, the challenge resides on how to account for the low structural similarity of different events. To combat this issue, we design a graph signature buffer to calculate the structural similarities of several streaming events and adjust the training of the global model accordingly. We evaluate the proposed model on the link weight prediction task on three real-world datasets of live video streaming events. Our experiments demonstrate the effectiveness of our proposed model, with an average relative gain of 25% against state-of-the-art strategies. For reproduction purposes, our evaluation datasets and implementation are publicly available at https://github.com/stefanosantaris/melanie

In this paper we present a deep graph reinforcement learning model to predict and improve the user experience during a live video streaming event, orchestrated by an agent/tracker. We first formulate the user experience prediction problem as a classification task, accounting for the fact that most of the viewers at the beginning of an event have poor quality of experience due to low-bandwidth connections and limited interactions with the tracker. In our model we consider different factors that influence the quality of user experience and train the proposed model on diverse state-action transitions when viewers interact with the tracker. In addition, provided that past events have various user experience characteristics we follow a gradient boosting strategy to compute a global model that learns from different events. Our experiments with three real-world datasets of live video streaming events demonstrate the superiority of the proposed model against several baseline strategies. Moreover, as the majority of the viewers at the beginning of an event has poor experience, we show that our model can significantly increase the number of viewers with high quality experience by at least 75% over the first streaming minutes. Our evaluation datasets and implementation are publicly available at https://publicresearch.z13.web.core.windows.net © 2021 IEEE.

Accounting for the fact that users have different sequential patterns, the main drawback of state-of-the-art recommendation strategies is that a fixed sequence length of user-item interactions is required as input to train the models. This might limit the recommendation accuracy, as in practice users follow different trends on the sequential recommendations. Hence, baseline strategies might ignore important sequential interactions or add noise to the models with redundant interactions, depending on the variety of users’ sequential behaviours. To overcome this problem, in this study we propose the SAR model, which not only learns the sequential patterns but also adjusts the sequence length of user-item interactions in a personalized manner. We first design an actor-critic framework, where the RL agent tries to compute the optimal sequence length as an action, given the user’s state representation at a certain time step. In addition, we optimize a joint loss function to align the accuracy of the sequential recommendations with the expected cumulative rewards of the critic network, while at the same time we adapt the sequence length with the actor network in a personalized manner. Our experimental evaluation on four real-world datasets demonstrates the superiority of our proposed model over several baseline approaches. Finally, we make our implementation publicly available at https://github.com/stefanosantaris/sar.

Dynamic graph representation learning strategies are based on different neural architectures to capture the graph evolution over time. However, the underlying neural architectures require a large amount of parameters to train and suffer from high online inference latency, that is several model parameters have to be updated when new data arrive online. In this study we propose Distill2Vec, a knowledge distillation strategy to train a compact model with a low number of trainable parameters, so as to reduce the latency of online inference and maintain the model accuracy high. We design a distillation loss function based on Kullback-Leibler divergence to transfer the acquired knowledge from a teacher model trained on offline data, to a small-size student model for online data. Our experiments with publicly available datasets show the superiority of our proposed model over several state-of-the-art approaches with relative gains up to 5% in the link prediction task. In addition, we demonstrate the effectiveness of our knowledge distillation strategy, in terms of number of required parameters, where Distill2Vec achieves a compression ratio up to 7:100 when compared with baseline approaches. For reproduction purposes, our implementation is publicly available at https://stefanosantaris.github.io/Distill2Vec.

Live video streaming has become a mainstay as a standard communication solution for several enterprises worldwide. To efficiently stream high-quality live video content to a large amount of offices, companies employ distributed video streaming solutions which rely on prior knowledge of the underlying evolving enterprise network. However, such networks are highly complex and dynamic. Hence, to optimally coordinate the live video distribution, the available network capacity between viewers has to be accurately predicted. In this paper we propose a graph representation learning technique on weighted and dynamic graphs to predict the network capacity, that is the weights of connections/links between viewers/nodes. We propose VStreamDRLS, a graph neural network architecture with a self-attention mechanism to capture the evolution of the graph structure of live video streaming events. VStreamDRLS employs the graph convolutional network (GCN) model over the duration of a live video streaming event and introduces a self-attention mechanism to evolve the GCN parameters. In doing so, our model focuses on the GCN weights that are relevant to the evolution of the graph and generate the node representation, accordingly. We evaluate our proposed approach on the link prediction task on two real-world datasets, generated by enterprise live video streaming events. The duration of each event lasted an hour. The experimental results demonstrate the effectiveness of VStreamDRLS when compared with state-of-the-art strategies. Our evaluation datasets and implementation are publicly available at https://github.com/stefanosantaris/vstreamdrls.

In this study, we present a dynamic graph representation learning model on weighted graphs to accurately predict the network capacity of connections between viewers in a live video streaming event. We propose EGAD, a neural network architecture to capture the graph evolution by introducing a self-attention mechanism on the weights between consecutive graph convolutional networks. In addition, we account for the fact that neural architectures require a huge amount of parameters to train, thus increasing the online inference latency and negatively influencing the user experience in a live video streaming event. To address the problem of the high online inference of a vast number of parameters, we propose a knowledge distillation strategy. In particular, we design a distillation loss function, aiming to first pretrain a teacher model on offline data, and then transfer the knowledge from the teacher to a smaller student model with less parameters. We evaluate our proposed model on the link prediction task on three real-world datasets, generated by live video streaming events. The events lasted 80 minutes and each viewer exploited the distribution solution provided by the company Hive Streaming AB. The experiments demonstrate the effectiveness of the proposed model in terms of link prediction accuracy and number of required parameters, when evaluated against state-of-the-art approaches. In addition, we study the distillation performance of the proposed model in terms of compression ratio for different distillation strategies, where we show that the proposed model can achieve a compression ratio up to 15:100, preserving high link prediction accuracy. For reproduction purposes, our evaluation datasets and implementation are publicly available at https://stefanosantaris.github.io/EGAD.

Publish/subscribe (pub/sub) mechanisms constitute an attractive communication paradigm in the design of large-scale notification systems for Online Social Networks (OSNs). To accommodate the large-scale workloads of notifications produced by OSNs, pub/sub mechanisms require thousands of servers distributed on different data centers all over the world, incurring large overheads. To eliminate the pub/sub resources used, we propose SELECT - a distributed pub/sub social notification system over peer-to-peer (P2P) networks. SELECT organizes the peers on a ring topology and provides an adaptive P2P connection establishment algorithm where each peer identifies the number of connections required, based on the social structure and user availability. This allows to propagate messages to the social friends of the users using a reduced number of hops. The presented algorithm is an efficient heuristic to an NP-hard problem which maps workload graphs to structured P2P overlays inducing overall, close to theoretical, minimal number of messages. Experiments show that SELECT reduces the number of relay nodes up to 89% versus the state-of-the-art pub/sub notification systems. Additionally, we demonstrate the advantage of SELECT against socially-aware P2P overlay networks and show that the communication between two socially connected peers is reduced on average by at least 64% hops, while achieving 100% communication availability even under high churn.

In this article, a media storm indexing mechanism is presented, where media storms are defined as fast incoming batches. We propose an approximate media storm indexing mechanism to index/store massive image collections with varying incoming image rate. To evaluate the proposed indexing mechanism, two architectures are used: i) a baseline architecture, which utilizes a disk-based processing strategy and ii) an in-memory architecture, which uses the Flink distributed stream processing framework. This study is the first in the literature to utilize an in-memory processing strategy to provide a media storm indexing mechanism. In the experimental evaluation conducted on two image datasets, among the largest publicly available with 80 M and 1 B images, a media storm generator is implemented to evaluate the proposed media storm indexing mechanism on different indexing workloads, that is, images that come with high volume and different velocity at the scale of 105and 106incoming images per second. Using the approximate media storm indexing mechanism a significant speedup factor, equal to 26.32 on average, is achieved compared with conventional indexing techniques, while maintaining high search accuracy, after having indexed the media storms. Finally, the implementations of both architectures and media storm indexing mechanisms are made publicly available.

We propose a media storm indexing algorithm using Map-Reduce in our recently proposed CDVC framework. In this study, CDVC is built on Flink, an open-source platform for stream data processing. The question we answer is how to store massive image collections; for instance, with over one million images per second, as well as with varying incoming rate. In our experiments with two benchmark datasets of 80M and 1B image descriptors, we evaluate the proposed algorithm on different indexing workloads, that is, images that come with high volume and different velocity at the scale of 105-106 images per second. Using a limited set of computational nodes, we show that we achieve a significant speed up factor of nine, on average, compared to conventional indexing techniques, in all settings. Finally, we make our source code publicly available.

In recognition that in modern applications billions of images are stored into distributed databases in different logical or physical locations, we propose a similarity search strategy over the cloud based on the dimensions value cardinalities of image descriptors. Our strategy has low preprocessing requirements by dividing the computational cost of the preprocessing steps into several nodes over the cloud and locating the descriptors with similar dimensions value cardinalities logically close. New images are inserted into the distributed databases over the cloud efficiently, by supporting dynamical update in real-time. The proposed insertion algorithm has low computational complexity, depending exclusively on the dimensionality of descriptors and a small subset of descriptors with similar dimensions value cardinalities. Finally, an efficient query processing algorithm is proposed, where the dimensions of image descriptors are prioritized in the searching strategy, assuming that dimensions of high value cardinalities have more discriminative power than the dimensions of low ones. The computation effort of the query processing algorithm is divided into several nodes over the cloud infrastructure. In our experiments with seven publicly available datasets of image descriptors, we show that the proposed similarity search strategy outperforms competitive methods of single node, parallel and cloud-based architectures, in terms of preprocessing cost, search time and accuracy.

Today's web commercial applications demand more powerful recommendation systems due to the rapid increase in the number of both consumers and available products. Searching for the best algorithm with the highest accuracy and realistic complexity is, most of the time, a very costly process in terms of both time and resources. In this paper we suggest an alternative framework called Hydra which enables the virtual fusion of any and as many currently available recommendation algorithms in such a distributed manner that algorithms' complexities are not summarized but parallelized. Therefore, we utilize the available algorithms and technologies aiming to achieve better accuracy in order to surpass even the most state of the art algorithms. In addition, Hydra can be used to find how algorithms interact with each other in order to estimate the resulting accuracy towards inventing a more precise algorithm diminishing the risk of a failed investment. Hydra can be adjusted and integrated in any recommendation application while it is also open to new functionalities which can be embedded easily and in a transparent manner.