Toon De Pessemier

Context aware recommendations for user-generated content on a social network site

The enormous offer of video content on the internet and its continuous growth make the selection process increasingly difficult for end-users. This overabundance of audio-visual material can be handled by a recommendation system that observes user preferences and assists people with finding interesting content. However, present-day recommendation systems focus on the metadata or the previous consumption behaviour to select the content but do not consider contextual information or social network relations. Therefore, we developed a tag cloud based recommendation system for user-generated content which exploits these social network relations. Recommendations based on the users profile are supplemented with social recommendations: content suggestions from people on the users contact list. Moreover, since we believe that the consumption context (location, time, etc.) has a significant influence on the content selection process, the system records all the available context information. Our next task is to analyze the obtained dataset and to determine the influence of the individual context features on the consumption behaviour. The system recommendations and social recommendations will be compared on the basis of effectiveness, novelty and user appreciation. Finally, we intend to incorporate the results of this analysis in our personalization algorithm in order to improve the recommendation results.

Context-aware recommendations through context and activity recognition in a mobile environment

The mobile Internet introduces new opportunities to gain insight in the user’s environment, behavior, and activity. This contextual information can be used as an additional information source to improve traditional recommendation algorithms. This paper describes a framework to detect the current context and activity of the user by analyzing data retrieved from different sensors available on mobile devices. The framework can easily be extended to detect custom activities and is built in a generic way to ensure easy integration with other applications. On top of this framework, a recommender system is built to provide users a personalized content offer, consisting of relevant information such as points-of-interest, train schedules, and touristic info, based on the user’s current context. An evaluation of the recommender system and the underlying context recognition framework shows that power consumption and data traffic is still within an acceptable range. Users who tested the recommender system via the mobile application confirmed the usability and liked to use it. The recommendations are assessed as effective and help them to discover new places and interesting information.

Comparison of group recommendation algorithms

In recent years recommender systems have become the common tool to handle the information overload problem of educational and informative web sites, content delivery systems, and online shops. Although most recommender systems make suggestions for individual users, in many circumstances the selected items (e.g., movies) are not intended for personal usage but rather for consumption in groups. This paper investigates how effective group recommendations for movies can be generated by combining the group members’ preferences (as expressed by ratings) or by combining the group members’ recommendations. These two grouping strategies, which convert traditional recommendation algorithms into group recommendation algorithms, are combined with five commonly used recommendation algorithms to calculate group recommendations for different group compositions. The group recommendations are not only assessed in terms of accuracy, but also in terms of other qualitative aspects that are important for users such as diversity, coverage, and serendipity. In addition, the paper discusses the influence of the size and composition of the group on the quality of the recommendations. The results show that the grouping strategy which produces the most accurate results depends on the algorithm that is used for generating individual recommendations. Therefore, the paper proposes a combination of grouping strategies which outperforms each individual strategy in terms of accuracy. Besides, the results show that the accuracy of the group recommendations increases as the similarity between members of the group increases. Also the diversity, coverage, and serendipity of the group recommendations are to a large extent dependent on the used grouping strategy and recommendation algorithm. Consequently for (commercial) group recommender systems, the grouping strategy and algorithm have to be chosen carefully in order to optimize the desired quality metrics of the group recommendations. The conclusions of this paper can be used as guidelines for this selection process.

Proposed architecture and algorithm for personalized advertising on iDTV and mobile devices

The advances in digital media entail an increase in the number of interactive advertising channels: Internet, interactive digital television (iDTV) and mobile applications are gaining importance in the advertisement sector. New interactive advertisement formats can complement the traditional commercial breaks and increase the revenues. Personalization is a promising technique to reach the target audience precisely, get a greater response and increase the return on investment. In this article we present an architecture that offers personalized commercials for iDTV, internet, and mobile devices. By logging the user activity on the three platforms, the system constructs a detailed profile usable for commercial targeting. This personal profile, supplemented with the community behavior and the metadata about the commercials, makes up the data source for the personalization algorithm.

Collaborative recommendations with content-based filters for cultural activities via a scalable event distribution platform

Nowadays, most people have limited leisure time and the offer of (cultural) activities to spend this time is enormous. Consequently, picking the most appropriate events becomes increasingly difficult for end-users. This complexity of choice reinforces the necessity of filtering systems that assist users in finding and selecting relevant events. Whereas traditional filtering tools enable e.g. the use of keyword-based or filtered searches, innovative recommender systems draw on user ratings, preferences, and metadata describing the events. Existing collaborative recommendation techniques, developed for suggesting web-shop products or audio-visual content, have difficulties with sparse rating data and can not cope at all with event-specific restrictions like availability, time, and location. Moreover, aggregating, enriching, and distributing these events are additional requisites for an optimal communication channel. In this paper, we propose a highly-scalable event recommendation platform which considers event-specific characteristics. Personal suggestions are generated by an advanced collaborative filtering algorithm, which is more robust on sparse data by extending user profiles with presumable future consumptions. The events, which are described using an RDF/OWL representation of the EventsML-G2 standard, are categorized and enriched via smart indexing and open linked data sets. This metadata model enables additional content-based filters, which consider event-specific characteristics, on the recommendation list. The integration of these different functionalities is realized by a scalable and extendable bus architecture. Finally, focus group conversations were organized with external experts, cultural mediators, and potential end-users to evaluate the event distribution platform and investigate the possible added value of recommendations for cultural participation.

A user-centric evaluation of context-aware recommendations for a mobile news service

Traditional recommender systems provide personal suggestions based on the user’s preferences, without taking into account any additional contextual information, such as time or device type. The added value of contextual information for the recommendation process is highly dependent on the application domain, the type of contextual information, and variations in users’ usage behavior in different contextual situations. This paper investigates whether users utilize a mobile news service in different contextual situations and whether the context has an influence on their consumption behavior. Furthermore, the importance of context for the recommendation process is investigated by comparing the user satisfaction with recommendations based on an explicit static profile, content-based recommendations using the actual user behavior but ignoring the context, and context-aware content-based recommendations incorporating user behavior as well as context. Considering the recommendations based on the static profile as a reference condition, the results indicate a significant improvement for recommendations that are based on the actual user behavior. This improvement is due to the discrepancy between explicitly stated preferences (initial profile) and the actual consumption behavior of the user. The context-aware content-based recommendations did not significantly outperform the content-based recommendations in our user study. Context-aware content-based recommendations may induce a higher user satisfaction after a longer period of service operation, enabling the recommender to overcome the cold-start problem and distinguish user preferences in various contextual situations.

Design and evaluation of a group recommender system

Though most recommender systems make suggestions for individual users, in many circumstances the selected items (e.g., movies) are not for personal usage but rather for consumption in group. In this paper, we present a recommender system for audio-visual content that generates suggestions for groups of people (such as families or friends) in the home environment. In this context, different group recommendation strategies are evaluated for various algorithms and sizes of the group. An offline evaluation proves the assumption that for randomly composed groups the accuracy of all recommendation algorithms decreases if the group size grows. Besides, the results show that the group recommendation strategy which produces the most accurate results is depending on the algorithm that is used for generating individual recommendations. Consequently, if an existing recommender system for individuals is extended to a recommender system for groups, the group recommendation strategy has to be chosen based on the utilized recommendation algorithm in order to maximize the efficiency of the group recommendations.

Extending the Bayesian Classifier to a Context-Aware Recommender System for Mobile Devices

Mobile devices that are capable of playing Internet videos have become wide-spread in recent years. Because of the enormous offer of video content, the lack of sufficient presentation space on the screen, and the laborious navigation on mobile devices, the video consumption process becomes more complicated for the end-user. To handle this problem, people need new instruments to assist with the hunting, filtering and selection process. We developed a methodology for mobile devices that makes the huge content sources more manageable by creating a user profile and personalizing the offer. This paper reports the structure of the user profile, the user interaction mechanism, and the recommendation algorithm, an improved version of the Bayesian classifier that incorporates aspects of the consumption context (like time, location, and mood of the user) to make the suggestions more accurate.

Analysis of the quality of experience of a commercial voice-over-IP service

Voice-over-IP (VoIP) services, enabling users to make cheap phone calls using the Internet, are becoming increasingly popular, not only on desktop computers but also on mobile devices such as smartphones that are connected through mobile networks. Users’ perception of the level of quality plays a key role in making a VoIP service to succeed or to fail. This paper demonstrates the influence of technical parameters (such as the audio codec, type of data network, and handovers during the call), device characteristics (such as the platform, manufacturer, model, and operating system), and application aspects (such as the software version and configuration) on the subjective quality of a commercial VoIP service. The relative influence of all these parameters is determined and a decision tree combines these results in order to assess the subjective quality. Combining a large number of objective parameters in a decision tree to determine the user’s subjective evaluation of the quality of a VoIP call is a novel and complex procedure. The subjective quality, in turn, has an influence on the duration of the call, and as a result an influence on the usage behavior of the service. The users’ assessment of the service quality is not evaluated by merely taking a snapshot of the perceived quality at one moment in time but rather by analyzing the perceived quality over a longer period of time during service usage, which has not been done up to now. Analyzing the VoIP service using a regression analysis over a period of 120 days showed that the perceived quality decreases slightly when the user utilizes the service more often and gets more familiar with it.

Time dependency of data quality for collaborative filtering algorithms

The efficiency of personal suggestions generated by collaborative filtering techniques is highly dependent on the quality and quantity of the available consumption data. Extending data sets with additional consumption data (from the past) might enrich the user profiles and generally leads to more accurate recommendations. Although if a considerable amount of profile information is already available and detailed personal preferences can be derived, supplementary consumption data may not have any (or a very limited) added value for the recommendation algorithm. These additional consumption data increase the required storage capacity and the computational load to generate the personal recommendations. Moreover, since personal preferences and the relevance of content items may vary over time, older consumption data might be outdated and lead to inaccurate recommendations. Therefore, we investigate which consumption data are (the most) relevant to feed the conventional collaborative filtering algorithms. For provider-generated content systems, we demonstrate that the accuracy of collaborative filtering algorithms increases by extending user profiles with additional older consumption data. In contrast, we witness the opposite effect for user-generated content systems: involving older consumption data has a negative influence on the recommender accuracy. These results are important for website owners who intend to employ a recommendation system at a minimum storage and computation cost.