Bruno Veloso

A Multi-Agent Brokerage Platform for Media Content Recommendation

Abstract Near real time media content personalisation is nowadays a major challenge involving media content sources, distributors and viewers. This paper describes an approach to seamless recommendation, negotiation and transaction of personalised media content. It adopts an integrated view of the problem by proposing, on the business-to-business (B2B) side, a brokerage platform to negotiate the media items on behalf of the media content distributors and sources, providing viewers, on the business-to-consumer (B2C) side, with a personalised electronic programme guide (EPG) containing the set of recommended items after negotiation. In this setup, when a viewer connects, the distributor looks up and invites sources to negotiate the contents of the viewer personal EPG. The proposed multi-agent brokerage platform is structured in four layers, modelling the registration, service agreement, partner lookup, invitation as well as item recommendation, negotiation and transaction stages of the B2B processes. The recommendation service is a rule-based switch hybrid filter, including six collaborative and two content-based filters. The rule-based system selects, at runtime, the filter(s) to apply as well as the final set of recommendations to present. The filter selection is based on the data available, ranging from the history of items watched to the ratings and/or tags assigned to the items by the viewer. Additionally, this module implements (i) a novel item stereotype to represent newly arrived items, (ii) a standard user stereotype for new users, (iii) a novel passive user tag cloud stereotype for socially passive users, and (iv) a new content-based filter named the collinearity and proximity similarity (CPS). At the end of the paper, we present off-line results and a case study describing how the recommendation service works. The proposed system provides, to our knowledge, an excellent holistic solution to the problem of recommending multimedia contents.

Personalised Advertising Supported by Agents

This paper reports the development of a B2B platform for the personalization of the publicity transmitted during the program intervals. The platform as a whole must ensure that the intervals are filled with ads compatible with the profile, context and expressed interests of the viewers. The platform acts as an electronic marketplace for advertising agencies (content producer companies) and multimedia content providers (content distribution companies). The companies, once registered at the platform, are represented by agents who negotiate automatically the price of the interval timeslots according to the specified price range and adaptation behaviour. The candidate ads for a given viewer interval are selected through a matching mechanism between ad, viewer and the current context (program being watched) profiles. The overall architecture of the platform consists of a multiagent system organized into three layers consisting of: (i) interface agents that interact with companies; (ii) enterprise agents that model the companies, and (iii) delegate agents that negotiate a specific ad or interval. The negotiation follows a variant of the Iterated Contract Net Interaction Protocol (ICNIP) and is based on the price/s offered by the advertising agencies to occupy the viewer’s interval.

Personalised fading for stream data

This paper describes a forgetting technique for the live update of viewer profiles based on individual sliding windows, fading and incremental matrix factorization. The individual sliding window maintains, for each viewer, a queue holding the last n viewer ratings. As new viewer events occur, they are inserted in the viewer queue, by shifting and fading the queue ratings, and the viewer latent model is faded. We explored time, rating-and-position and popularity-based fading techniques, using the latter as the base fading algorithm. This approach attempts to address the problem of dynamic viewer profile updating (volatile preferences) as well as the problem of bounded processing resources (fixed size queues). The results show that our approach outperforms previous approaches, improving the quality of the predictions.

Media Brokerage: Agent-Based SLA Negotiation

Media content personalisation is a major challenge involving viewers as well as media content producer and distributor businesses. The goal is to provide viewers with media items aligned with their interests. Producers and distributors engage in item negotiations to establish the corresponding service level agreements (SLA). In order to address automated partner lookup and item SLA negotiation, this paper proposes the MultiMedia Brokerage (MMB) platform, which is a multiagent system that negotiates SLA regarding media items on behalf of media content producer and distributor businesses. The MMB platform is structured in four service layers: interface, agreement management, business modelling and market. In this context, there are: (i) brokerage SLA (bSLA), which are established between individual businesses and the platform regarding the provision of brokerage services; and (ii) item SLA (iSLA), which are established between producer and distributor businesses about the provision of media items. In particular, this paper describes the negotiation, establishment and enforcement of bSLA and iSLA, which occurs at the agreement and negotiation layers, respectively. The platform adopts a pay-per-use business model where the bSLA define the general conditions that apply to the related iSLA. To illustrate this process, we present a case study describing the negotiation of a bSLA instance and several related iSLA instances. The latter correspond to the negotiation of the Electronic Program Guide (EPG) for a specific end viewer.

CloudAnchor: Agent-Based Brokerage of Federated Cloud Resources

This paper presents CloudAnchor, a brokerage platform conceived to help Small and Medium Sized Enterprises (SME) embrace Infrastructure as a Service (IaaS) cloud computing both as providers and consumers. The platform, which transacts automatically single and federated IaaS cloud resources, is a multi-layered Multi-Agent System (MAS) where providers, consumers and virtual providers, representing provider coalitions, are modelled by dedicated agents. Federated resources are detained and negotiated by virtual providers on behalf of the corresponding coalition of providers. CloudAnchor negotiates and establishes Service Level Agreements (SLA) on behalf of SME businesses regarding the provision of brokerage services as well as the provision of single and federated IaaS resources. The discovery, invitation, acceptance and negotiation processes rely on a distributed trust model designed to select the best business partners for consumers and providers and improve runtime.

Personalised Dynamic Viewer Profiling for Streamed Data

Nowadays, not only the number of multimedia resources available is increasing exponentially, but also the crowd-sourced feedback volunteered by viewers generates huge volumes of ratings, likes, shares and posts/reviews. Since the data size involved surpasses human filtering and searching capabilities, there is the need to create and maintain the profiles of viewers and resources to develop recommendation systems to match viewers with resources. In this paper, we propose a personalised viewer profiling technique which creates individual viewer models dynamically. This technique is based on a novel incremental learning algorithm designed for stream data. The results show that our approach outperforms previous approaches, reducing substantially the prediction errors and, thus, increasing the accuracy of the recommendations.

APASail—An Agent-Based Platform for Autonomous Sailing Research and Competition

This paper presents an Agent-based Platform for Autonomous Sailing—APASail—intended to support real and simulated competitions as well as assess navigation and boat modelling algorithms. The platform provides back-end services—competition server, boat modelling and data storage—and supports external browsers and software agents as front-end clients. The back-end adopts the Multi-Agent System (MAS) paradigm for the internal modelling of sailing boats and offers a Web Service Application Programming Interface (API) for the external software agents and a Web application for Web browsers. As a whole, the platform offers tracking (real competitions) and simulation (simulated competitions) modes. The testing and assessment of navigation algorithms and boat models correspond to private simulated competitions. In simulation mode, the back-end internal boat agent implements a simplified physical model, including the weight, sail area, angle of the sail and rudder, velocity and direction of the wind and position and velocity of the hull, whereas the front-end external boat agent implements the navigation algorithm on the team side, ensuring the privacy of strategic knowledge. The Web application allows the configuration and launching of competitions, the registration of teams and researchers, the uploading of boat physical features for simulation as well as the live or playback viewing of real and simulated competitions. The simulation mode is illustrated with the help of a case study. The APASail platform, which is open, scalable, modular and distributed, was designed for the research community to prepare, run and gather data from real and simulated autonomous sailing competitions.

Self Hyper-Parameter Tuning for Data Streams

The widespread usage of smart devices and sensors together with the ubiquity of the Internet access is behind the exponential growth of data streams. Nowadays, there are hundreds of machine learning algorithms able to process high-speed data streams. However, these algorithms rely on human expertise to perform complex processing tasks like hyper-parameter tuning. This paper addresses the problem of data variability modelling in data streams. Specifically, we propose and evaluate a new parameter tuning algorithm called Self Parameter Tuning (SPT). SPT consists of an online adaptation of the Nelder & Mead optimisation algorithm for hyper-parameter tuning. The method explores a dynamic size sample method to evaluate the current solution, and uses the Nelder & Mead operators to update the current set of parameters. The main contribution is the adaptation of the Nelder-Mead algorithm to automatically tune regression hyper-parameters for data streams. Additionally, whenever concept drifts occur in the data stream, it re-initiates the search for new hyper-parameters. The proposed method has been evaluated on regression scenario. Experiments with well known time-evolving data streams show that the proposed SPT hyper-parameter optimisation outperforms the results of previous expert hyper-parameter tuning efforts.

Renegotiation of Electronic Brokerage Contracts

CloudAnchor is a multiagent e-commerce platform which offers brokerage and resource trading services to Infrastructure as a Service (IaaS) providers and consumers. The access to these services requires the prior negotiation of Service Level Agreements (SLA) between the parties. In particular, the brokerage SLA (bSLA), which is mandatory for a business to have access to the platform, specifies the brokerage fee the business will pay every time it successfully trades a resource within the platform. However, while the negotiation of the resource SLA (rSLA) includes the uptime of the service, the brokerage SLA was negotiated for an unspecified time span. Since the commercial relationship – defined through the bSLA – between a business and the platform can be long lasting, it is essential for businesses to be able to renegotiate the bSLA terms, i.e., renegotiate the brokerage fee. To address this issue, we designed a bSLA renegotiation mechanism, which takes into account the duration of the bSLA as well as the past behaviour (trust) and success (transactions) of the business in the CloudAnchor platform. The results show that the implemented bSLA renegotiation mechanism privileges, first, the most reliable businesses, and, then, those with higher volume of transactions, ensuring that the most reliable businesses get the best brokerage fees and resource prices. The proposed renegotiation mechanism promotes the fulfilment of SLA by all parties and increases the satisfaction of the trustworthy businesses in the CloudAnchor platform.

Collaborative Filtering with Semantic Neighbour Discovery

Nearest neighbour collaborative filtering (NNCF) algorithms are commonly used in multimedia recommender systems to suggest media items based on the ratings of users with similar preferences. However, the prediction accuracy of NNCF algorithms is affected by the reduced number of items – the subset of items co-rated by both users – typically used to determine the similarity between pairs of users. In this paper, we propose a different approach, which substantially enhances the accuracy of the neighbour selection process – a user-based CF (UbCF) with semantic neighbour discovery (SND). Our neighbour discovery methodology, which assesses pairs of users by taking into account all the items rated at least by one of the users instead of just the set of co-rated items, semantically enriches this enlarged set of items using linked data and, finally, applies the Collinearity and Proximity Similarity metric (CPS), which combines the cosine similarity with Chebyschev distance dissimilarity metric. We tested the proposed SND against the Pearson Correlation neighbour discovery algorithm off-line, using the HetRec data set, and the results show a clear improvement in terms of accuracy and execution time for the predicted recommendations.