Skip to main content
main-content
Top

About this book

After the initial enthusiastic initiatives and investments and the eventual bubble, el- tronic commerce (EC) has changed and evolved into a well-established and founded reality both from a technological point of view and from a scientific one. Nevert- less, together with its evolution, new challenges and topics have emerged as well as new questions have been raised related to many aspects of EC. Keeping in mind the experience and the tradition of the past editions of EC-Web, we tried, for its 10th edition, to introduce some meaningful innovations about the structure and the sci- tific organization of the conference. Our main target was to highlight the autonomous role of the different (sometimes heterogeneous) aspects of EC, without missing their interdisciplinary scope. This required the conference to be organized into four “mi- conferences," each for a relevant area of EC and equipped with a corresponding Area Chair. Both the submission and the review process took into account the organization into four tracks, namely: “Service-Oriented E-Commerce and Business Processes,” “Recommender Systems,” “E-Payment, Security and Trust” and “Electronic C- merce and Web 3. 0. ” Therefore, the focus of the conference was to cover aspects related to the theoretical foundation of EC, business processes as well as new - proaches exploiting recently emerged technologies and scenarios such as the Semantic Web, Web services, SOA architectures, mobile and ubiquitous computing, just to cite a few.

Table of Contents

Frontmatter

Invited Talk

Voting: A View through the Algorithmic Lens

In recent years, voting and elections have been actively studied by computer scientists, both because of their interesting algorithmic properties and due to their potential applications to the design of multiagent systems and e-commerce. In this talk, I will give an overview of several recent developments in the nascent area of computational social choice and discuss their relevance to the use of voting in practical e-commerce applications.

Edith Elkind

Infomobility and Negotiation

Personalized Popular Blog Recommender Service for Mobile Applications

Weblogs have emerged as a new communication and publication medium on the Internet for diffusing the latest useful information. Providing value-added mobile services such as blog articles is increasingly important to attract mobile users to mobile commerce. There are, however, a tremendous number of blog articles, and mobile users generally have difficulty in browsing weblogs. Accordingly, providing mobile users with blog articles that suit their interests is an important issue. Very little research, however, focuses on this issue. In this work, we propose a Customized Content Service on a mobile device (m-CCS) to filter and push blog articles to mobile users. The m-CCS can predict the latest popular blog topics by forecasting the trend of time-sensitive popularity of weblogs. Furthermore, to meet the diversified interest of mobile users, m-CCS further analyzes users’ browsing logs to derive their interests, which are then used to recommend their preferred popular blog topics and articles. The prototype system of m-CCS demonstrates that the system can effectively recommend mobile users desirable blog articles with respect to both popularity and personal interests.

Pei-Yun Tsai, Duen-Ren Liu

Bargaining Agents in Wireless Contexts: An Alternating-Offers Protocol for Multi-issue Bilateral Negotiation in Mobile Marketplaces

We present an approach to multi-issue bilateral negotiation for mobile commerce scenarios. The negotiation mechanism has been integrated in a semantic-based application layer enhancing both RFID and Bluetooth wireless standards. OWL DL has been used to model advertisements and relationships among issues within a shared common ontology. Finally, non standard inference services integrated with utility theory help in finding suitable agreements. We illustrate and motivate the provided theoretical framework in a wireless commerce case study.

Azzurra Ragone, Michele Ruta, Eugenio Di Sciascio, Francesco M. Donini

A Group Recommender System for Tourist Activities

This paper introduces a method for giving recommendations of tourist activities to a group of users. This method makes recommendations based on the group tastes, their demographic classification and the places visited by the users in former trips. The group recommendation is computed from individual personal recommendations through the use of techniques such as aggregation, intersection or incremental intersection. This method is implemented as an extension of the

e-Tourism

tool, which is a user-adapted tourism and leisure application, whose main component is the

Generalist Recommender System Kernel (GRSK)

, a domain-independent taxonomy-driven search engine that manages the group recommendation.

Inma Garcia, Laura Sebastia, Eva Onaindia, Cesar Guzman

Personalized Location-Based Recommendation Services for Tour Planning in Mobile Tourism Applications

Travel and tour planning is a process of searching, selecting, grouping and sequencing destination related products and services including attractions, accommodations, restaurants, and activities. Personalized recommendation services aim at suggesting products and services to meet users’ preferences and needs, while location-based services focus on providing information based on users’ current positions. Due to the fast growing of user needs in the mobile tourism domain, how to provide personalized location-based tour recommendation services becomes a critical research and practical issue. The objective of this paper is to propose a system architecture and design methods for facilitating the delivery of location-based recommendation services to support personalized tour planning. Based on tourists’ current location and time, as well as personal preferences and needs, various recommendations regarding sightseeing spots, hotels, restaurants, and packaged tour plans can be generated efficiently. An application prototype is also implemented to illustrate and test the system feasibility and effectiveness.

Chien-Chih Yu, Hsiao-ping Chang

E-payments and Trust

Do You Trust Your Phone?

Despite the promising start, Electronic Commerce has not taken off mostly because of security issues with the communication infrastructures that are popping up threateningly undermining the perceived trustworthiness in Electronic Commerce.

Some Internet security issues, like malware, phishing, pharming are well known to the Internet community. Such issues are being, however, transferred to the telephone networks thanks to the symbiotic relation between the two worlds. Such an interconnection is becoming so pervasive that we can really start thinking about a unique network, which, in this paper, we refer to as the Interphonet.

The main goal of this paper is to analyze some of the Internet security issues that are being transferred to the Interphonet and also to identify new security issues of the Interphonet. In particular we will discuss about mobile phones malware and identity theft, phishing with SMS, telephone pharming, untraceability of phone calls that use VoIP and Caller ID spoofing. We will also briefly discuss about countermeasures.

Aniello Castiglione, Roberto De Prisco, Alfredo De Santis

A Multi-scheme and Multi-channel Framework for Micropayment Systems

In this paper we present an integrated framework for developing and running micropayment services. Our framework is multi-channel, as it allows a micropayment service to be used, at the same time, by clients using different types of communication channels. It is also multi-scheme, because it allows to have on the same server different types of micropayment schemes.

The framework has been designed in such a way to simplify the distribution and the replication of its server components across several machines, thus increasing the overall efficiency. On the client side, it includes two library of classes that can be used to run micropayment services on Java applications running on a desktop computer, in a Web browser or on a mobile phone. The framework also includes the implementation of two traditional micropayment schemes, as well as the communication modules needed to implement micropayment schemes over HTTP based and SMS based communication channels.

Aniello Castiglione, Giuseppe Cattaneo, Maurizio Cembalo, Pompeo Faruolo, Umberto Ferraro Petrillo

Secure Transaction Protocol for CEPS Compliant EPS in Limited Connectivity Environment

Common Electronic Purse Specification (CEPS) used by European countries, elaborately defines the transaction between customer’s CEP card and merchant’s point of sales (POS) terminal. However it merely defines the specification to transfer the transactions between the Merchant and Merchant Acquirer (MA). This paper proposes a novel approach by introducing an entity, mobile merchant acquirer (MMA) which is a trusted agent of MA and principally works on man in middle concept, but facilitates remote two fold mutual authentication and secure transaction transfer between Merchant and MA through MMA. This approach removes the bottle-neck of connectivity issues between Merchant and MA in limited connectivity environment. The proposed protocol ensures the confidentiality, integrity and money atomicity of transaction batch. The proposed protocol has been verified for correctness by Spin, a model checker and security properties of the protocol have been verified by avispa.

Satish Devane, Deepak Phatak

Trust Enhanced Authorization for Mobile Agents

Trust has been recognized as an important aspect for mobile agent security. In this paper, we develop a logic based trust model which enables the capturing of a comprehensive set of trust relationships to enhance the security of conventional access control mechanisms in a mobile based applications. We first discuss the notion of trust and its relevance to mobile agent security. Next we define a logic program based language to facilitate the modelling process. To enforce the security related trustworthy behaviours, we then define a set of general rules to capture the semantics. Finally, the language is applied in a mobile agent context to demonstrate how the trust can be explicitly modelled and reasoned about to support better security decisions for the mobile agent based systems.

Chun Ruan, Vijay Varadharajan

Domain Knowledge and Metadata Exploitation

Towards Semantic Modelling of Business Processes for Networked Enterprises

The paper presents an approach to the semantic modelling and annotation of business processes and information resources, as it was designed within the FP7 ICT EU project SPIKE to support creation and maintenance of short-term business alliances and networked enterprises. A methodology for the development of the resource ontology, as a shareable knowledge model for semantic description of business processes, is proposed. Systematically collected user requirements, conceptual models implied by the selected implementation platform as well as available ontology resources and standards are employed in the ontology creation. The process of semantic annotation is described and illustrated using an example taken from a real application case.

Karol Furdík, Marián Mach, Tomáš Sabol

Metadata-Driven SOA-Based Application for Facilitation of Real-Time Data Warehousing

Service-oriented architecture (SOA) has already been widely recognized as an effective paradigm for achieving integration of diverse information systems. SOA-based applications can cross boundaries of platforms, operation systems and proprietary data standards, commonly through the usage of Web Services technology. On the other side, metadata is also commonly referred to as a potential integration tool given the fact that standardized metadata objects can provide useful information about specifics of unknown information systems with which one has interest in communicating with, using an approach commonly called "model-based integration". This paper presents the result of research regarding possible synergy between those two integration facilitators. This is accomplished with a vertical example of a metadata-driven SOA-based business process that provides ETL (Extraction, Transformation and Loading) and metadata services to a data warehousing system in need of a real-time ETL support.

Damir Pintar, Mihaela Vranić, Zoran Skočir

Exploiting Domain Knowledge by Automated Taxonomy Generation in Recommender Systems

The effectiveness of incorporating domain knowledge into recommender systems to address their sparseness problem and improve their prediction accuracy has been discussed in many research works. However, this technique is usually restrained in practice because of its high computational expense. Although cluster analysis can alleviate the computational complexity of the recommendation procedure, it is not satisfactory in preserving pair-wise item similarities, which would severely impair the recommendation quality. In this paper, we propose an efficient approach based on the technique of Automated Taxonomy Generation to exploit relational domain knowledge in recommender systems so as to achieve high system scalability and prediction accuracy. Based on the domain knowledge, a hierarchical data model is synthesized in an offline phase to preserve the original pairwise item similarities. The model is then used by online recommender systems to facilitate the similarity calculation and keep their recommendation quality comparable to those systems by means of real-time exploiting domain knowledge. Experiments were conducted upon real datasets to evaluate our approach.

Tao Li, Sarabjot S. Anand

Automatic Generation of Mashups for Personalized Commerce in Digital TV by Semantic Reasoning

The evolution of information technologies is consolidating

recommender systems

as essential tools in e-commerce. To date, these systems have focused on discovering the items that best match the preferences, interests and needs of individual users, to end up listing those items by decreasing relevance in some menus. In this paper, we propose extending the current scope of recommender systems to better support trading activities, by automatically generating interactive applications that provide the users with personalized commercial functionalities related to the selected items. We explore this idea in the context of Digital TV advertising, with a system that brings together semantic reasoning techniques and new architectural solutions for web services and mashups.

Yolanda Blanco-Fernández, Martín López-Nores, José J. Pazos-Arias, Manuela I. Martín-Vicente

Invited Talk

Product Variety, Consumer Preferences, and Web Technology: Can the Web of Data Reduce Price Competition and Increase Customer Satisfaction?

E-Commerce on the basis of current Web technology has created fierce competition with a strong focus on price. Despite a huge variety of offerings and diversity in the individual preferences of consumers, current Web search fosters a very early reduction of the search space to just a few commodity makes and models. As soon as this reduction has taken place, search is reduced to flat price comparison. This is unfortunate for the manufacturers and vendors, because their individual value proposition for a particular customer may get lost in the course of communication over the Web, and it is unfortunate for the customer, because he/she may not get the most utility for the money based on her/his preference function. A key limitation is that consumers cannot search using a consolidated view on all alternative offers across the Web. In this talk, I will (1) analyze the technical effects of products and services search on the Web that cause this mismatch between supply and demand, (2) evaluate how the GoodRelations vocabulary and the current Web of Data movement can improve the situation, (3) give a brief hands-on demonstration, and (4) sketch business models for the various market participants.

Martin Hepp

Design and Modelling of Enterprise and Distributed Systems

Perspectives for Web Service Intermediaries: How Influence on Quality Makes the Difference

In the service-oriented computing paradigm and the Web service architecture, the broker role is a key facilitator to leverage technical capabilities of loose coupling to achieve organizational capabilities of dynamic customer-provider-relationships. In practice, this role has quickly evolved into a variety of intermediary concepts that refine and extend the basic functionality of service brokerage with respect to various forms of added value like platform or market mechanisms. While this has initially led to a rich variety of Web service intermediaries, many of these are now going through a phase of stagnation or even decline in customer acceptance. In this paper we present a comparative study on insufficient service quality that is arguably one of the key reasons for this phenomenon. In search of a differentiation with respect to quality monitoring and management patterns, we categorize intermediaries into Infomediaries, e-Hubs, e-Markets and Integrators. A mapping of quality factors and control mechanisms to these categories depicts their respective strengths and weaknesses. The results show that Integrators have the highest overall performance, followed by e-Markets, e-Hubs and lastly Infomediaries. A comparative market survey confirms the conceptual findings.

Ulrich Scholten, Robin Fischer, Christian Zirpins

Aligning Risk Management and Compliance Considerations with Business Process Development

The improvement of business processes, to date, primarily focuses on effectiveness and efficiency, thereby creating additional value for the organization and its stakeholders. The design of processes should also ensure that its result and the value obtained compensates for the risks affecting this value. In this paper the different kinds of risk affecting a business process are introduced, after which solutions to the problem of risk mitigation are discussed, resulting in a proposed framework to mollify these risks by incorporating a class of risk-mitigation rules into business process development.

Martijn Zoet, Richard Welke, Johan Versendaal, Pascal Ravesteyn

Electronic Commerce and Web 3.0

Using Knowledge Base for Event-Driven Scheduling of Web Monitoring Systems

Web monitoring systems report any changes to their target web pages by revisiting them frequently. As they operate under significant resource constraints, it is essential to minimize revisits while ensuring minimal delay and maximum coverage. Various statistical scheduling methods have been proposed to resolve this problem; however, they are static and cannot easily cope with events in the real world. This paper proposes a new scheduling method that manages unpredictable events. An MCRDR (Multiple Classification Ripple-Down Rules) document classification knowledge base was reused to detect events and to initiate a prompt web monitoring process independent of a static monitoring schedule. Our experiment demonstrates that the approach improves monitoring efficiency significantly.

Yang Sok Kim, Sung Won Kang, Byeong Ho Kang, Paul Compton

RCQ-GA: RDF Chain Query Optimization Using Genetic Algorithms

The application of Semantic Web technologies in an Electronic Commerce environment implies a need for good support tools. Fast query engines are needed for efficient querying of large amounts of data, usually represented using RDF. We focus on optimizing a special class of SPARQL queries, the so-called RDF chain queries. For this purpose, we devise a genetic algorithm called RCQ-GA that determines the order in which joins need to be performed for an efficient evaluation of RDF chain queries. The approach is benchmarked against a two-phase optimization algorithm, previously proposed in literature. The more complex a query is, the more RCQ-GA outperforms the benchmark in solution quality, execution time needed, and consistency of solution quality. When the algorithms are constrained by a time limit, the overall performance of RCQ-GA compared to the benchmark further improves.

Alexander Hogenboom, Viorel Milea, Flavius Frasincar, Uzay Kaymak

Integrating Markets to Bridge Supply and Demand for Knowledge Intensive Tasks

The advent of the knowledge-based economy has underlined the importance of intellectual capital that is possessed by knowledge intensive organizations. Three general observations of knowledge intensive work produced by actors working in such organizations served as the basis for the initiation of this research. First, knowledge intensive tasks become increasingly complex. Second, actors that perform such tasks experience an increase in cognitive load. Third, the desired quality of task performance and the produced task results are at stake due to the aforementioned two developments. In this research we investigate how supply and demand of intangible assets such as knowledge, cognitive characteristics, and quality factors can be matched based on market mechanisms.

Sietse Overbeek, Marijn Janssen, Patrick van Bommel

Real-Time Robust Adaptive Modeling and Scheduling for an Electronic Commerce Server

With the increasing importance and pervasiveness of Internet services, it is becoming a challenge for the proliferation of electronic commerce services to provide performance guarantees under extreme overload. This paper describes a real-time optimization modeling and scheduling approach for performance guarantee of electronic commerce servers. We show that an electronic commerce server may be simulated as a multi-tank system. A robust adaptive server model is subject to unknown additive load disturbances and uncertain model matching. Overload control techniques are based on adaptive admission control to achieve timing guarantees. We evaluate the performance of the model using a complex simulation that is subjected to varying model parameters and massive overload.

Bing Du, Chun Ruan

Collaboration-Based Approaches

Content-Based Personalization Services Integrating Folksonomies

Basic content-based personalization consists in matching up the attributes of a user profile, in which preferences and interests are stored, with the attributes of a content object. The Web 2.0 (r)evolution has changed the game for personalization, from ‘elitary’ Web 1.0, written by few and read by many, to web content generated by everyone (

user-generated content

- UGC), since the role of people has evolved from passive consumers of information to that of active contributors.

One of the forms of UGC that has drawn most attention of the research community is folksonomy, a taxonomy generated by users who collaboratively annotate and categorize resources of interests with freely chosen keywords called tags.

FIRSt (

F

olksonomy-based

I

tem

R

ecommender sy

St

em) is a content-based recommender system developed at the University of Bari which integrates UGC (through social tagging) in a classic content-based model, letting users express their preferences for items by entering a numerical rating as well as to annotate rated items with free tags. FIRSt is capable of providing recommendations for items in several domains (e.g., movies, music, books), provided that descriptions of items are available as text documents (e.g. plot summaries, reviews, short abstracts). This paper describes the system general architecture and user modeling approach, showing how this recommendation model has been applied to recommend the artworks located at the Vatican Picture Gallery (Pinacoteca Vaticana), providing users with a personalized museum tour tailored on their tastes.

Cataldo Musto, Fedelucio Narducci, Pasquale Lops, Marco de Gemmis, Giovanni Semeraro

Computational Complexity Reduction for Factorization-Based Collaborative Filtering Algorithms

Alternating least squares (ALS) is a powerful matrix factorization (MF) algorithm for both implicit and explicit feedback based recommender systems. We show that by using the Sherman-Morrison formula (SMF), we can reduce the computational complexity of several ALS based algorithms. It also reduces the complexity of greedy forward and backward feature selection algorithms by an order of magnitude. We propose linear kernel ridge regression (KRR) for users with few ratings. We show that both SMF and KRR can efficiently handle new ratings.

István Pilászy, Domonkos Tikk

Sequence-Based Trust for Document Recommendation

Collaborative Filtering (CF) recommender systems have emerged in various applications to support item recommendation, which solve the information-overload problem by suggesting items of interests to users. Recently, trust-based recommender systems have incorporated the trustworthiness of users into the CF techniques to improve the quality of recommendation. They propose trust computation models to derive the trust value based on users’ past ratings on items. A user is more trustworthy if he has contributed more accurate predictions than other users. Nevertheless, none of them derive the trust value based on a sequence of user’s ratings on items. We propose a sequence-based trust model to derive the trust value based on users’ sequences of ratings on documents. In knowledge-intensive environments, users normally have various information needs to access required documents over time, which forms a sequence of documents ordered according to their access time. The model considers two factors - time factor and document similarity in computing the trustworthiness of users. The proposed model is incorporated into standard collaborative filtering method to discover trustworthy neighbors for making predictions. The experiment result shows that the proposed model can improve the prediction accuracy of CF method comparing to other trust-based recommender systems.

Hsuan Chiu, Duen-Ren Liu, Chin-Hui Lai

Recommender Systems Modelling

Recommender Systems on the Web: A Model-Driven Approach

Recommendation techniques have been increasingly incorporated in e-commerce applications, supporting clients in identifying those items that best fit their needs. Unfortunately, little effort has been made to integrate these techniques into methodological proposals of Web development, discouraging the adoption of engineering approaches to face the complexity of recommender systems. This paper introduces a proposal to develop Web-based recommender systems from a model-driven perspective, specifying the elements of recommendation algorithms from a high abstraction level. Adopting the item-to-item approach, this proposal adopts the conceptual models of an existing Web development process to represent the preferences of users for different items, the similarity between obtained from different algorithms, and the selection and ordering of the recommended items according to a predicted rating value. Along with systematizing the development of these systems, this approach permits to evaluate different algorithms with minor changes at conceptual level, simplifying their mapping to final implementations.

Gonzalo Rojas, Francisco Domínguez, Stefano Salvatori

Designing a Metamodel-Based Recommender System

Current recommender systems have to cope with a certain reservation because they are considered to be hard to maintain and to give rather schematic advice. This paper presents an approach to increase maintainability by generating essential parts of the recommender system based on thorough metamodeling. Moreover, preferences are elicited on the basis of user needs rather than product features thus leading to a more user-oriented behavior. The metamodel-based design allows to efficiently adapt all domain-dependent parts of the system.

Sven Radde, Bettina Zach, Burkhard Freitag

Towards Privacy Compliant and Anytime Recommender Systems

Recommendation technologies have traditionally been used in domains such as E-commerce and Web navigation to recommend resources to customers so as to help them to get the pertinent resources. Among the possible approaches is collaborative filtering that does not take into account the content of the resources: only the traces of usage of the resources are considered. State of the art models, such as sequential association-rules and Markov models, that can be used in the frame of privacy concerns, are usually studied in terms of performance, state space complexity and time complexity. Many of them have a large time complexity and require a long time to compute recommendations. However, there are domains of application of the models where recommendations may be required quickly. This paper focuses on the study of how these state of the art models can be adapted so as to be anytime. In that case recommendations can be proposed to the user whatever is the computation time available, the quality of the recommendations increases according to the computation time. We show that such models can be adapted so as to be anytime and we propose several strategies to compute recommendations iteratively. We also show that the computation time needed by these new models is not increased compared to classical ones; even so, it sometimes decreases.

Armelle Brun, Anne Boyer

Reputation and Fraud Detection

Assessing Robustness of Reputation Systems Regarding Interdependent Manipulations

Reputation systems are subject to several types of manipulations, often in context of fraud. The current literature offers mainly partial solutions for specific manipulations. However, in practice a reputation system should be robust against all relevant threats. This paper explores the combination of several partial solutions in an evolutionary simulation model. The analysis shows that some partial solutions interfere with each other. In particular, it turns out that there is a crucial tradeoff between sanctioning and rehabilitation of bad behaviour that can be solved by a minimal transaction fee.

Ivo Reitzenstein, Ralf Peters

Fraud Detection by Human Agents: A Pilot Study

Fraud is a constant problem for online auction sites. Besides failures in detecting fraudsters, the currently employed methods yield many false positives: bona fide sellers that end up harassed by the auction site as suspects. We advocate the use of human computation (also called crowdsourcing) to improve precision and recall of current fraud detection techniques. To examine the feasibility of our proposal, we did a pilot study with a set of human subjects, testing whether they could distinguish fraudsters from common sellers before negative feedback arrived and looking just at a snapshot of seller profiles. Here we present the methodology used and the obtained results, in terms of precision and recall of human classifiers, showing positive evidence that detecting fraudsters with human computation is viable.

Vinicius Almendra, Daniel Schwabe

Recommender Systems and the Social Web

Finding My Needle in the Haystack: Effective Personalized Re-ranking of Search Results in Prospector

This paper provides an overview of Prospector, a personalized Internet meta-search engine, which utilizes a combination of ontological information, ratings-based models of user interests, and complementary theme-oriented group models to recommend (through re-ranking) search results obtained from an underlying search engine. Re-ranking brings “closer to the top” those items that are of particular interest to a user or have high relevance to a given theme. A user-based, real-world evaluation has shown that the system is effective in promoting results of interest, but lags behind Google in user acceptance, possibly due to the absence of features popularized by said search engine. Overall, users would consider employing a personalized search engine to perform searches with terms that require disambiguation and / or contextualization.

Florian König, Lex van Velsen, Alexandros Paramythis

RATC: A Robust Automated Tag Clustering Technique

Nowadays, the most dominant and noteworthy web information sources are developed according to the collaborative-web paradigm, also known as Web 2.0. In particular, it represents a novel paradigm in the way users interact with the web. Users (also called prosumers) are no longer passive consumers of published content, but become involved, implicitly and explicitly, as they cooperate by providing their own resources in an “architecture of participation”. In this scenario, collaborative tagging, i.e., the process of classifying shared resources by using keywords, becomes more and more popular. The main problem in such task is related to well-known linguistic phenomena, such as polysemy and synonymy, making effective content retrieval harder. In this paper, an approach that monitors users activity in a tagging system and dynamically quantifies associations among tags is presented. The associations are then used to create tags clusters. Experiments are performed comparing the proposed approach with a state-of-the-art tag clustering technique. Results –given in terms of classical precision and recall– show that the approach is quite effective in the presence of strongly related tags in a cluster.

Ludovico Boratto, Salvatore Carta, Eloisa Vargiu

Recommender Systems in Action

ISeller: A Flexible Personalization Infrastructure for e-Commerce Applications

We present

ISeller

, an industrial-strength recommendation system for a diverse range of commercial application domains. The system supports several recommendation paradigms such as collaborative, content-based and knowledge-based filtering, as well as one-shot and conversational interaction modes out of the box. A generic user modeling component allows different forms of hybrid personalization and enables the system to support process-oriented interactive selling in various product domains. This paper contributes a detailed discussion of a domain independent and flexible recommendation system from a software architecture viewpoint and illustrates it with different usage scenarios.

Markus Jessenitschnig, Markus Zanker

Comparing Pre-filtering and Post-filtering Approach in a Collaborative Contextual Recommender System: An Application to E-Commerce

Recent literature predicts that including context in a recommender system may improve its performance. The context-based recommendation approaches are classified as pre-filtering, post-filtering and contextual modeling. Little research has been done on studying whether including context in a recommender system improves the recommendation performance and no research has compared yet the different approaches to contextual RS. The research contribution of this work lies in studying the effect of the context on the recommendation performance and comparing a pre-filtering approach to a post-filtering using a collaborative filtering recommender system.

Umberto Panniello, Michele Gorgoglione, Cosimo Palmisano

Providing Relevant Background Information in Smart Environments

In this paper we describe a system, called GAIN (Group Adapted Interaction for News), which selects background information to be displayed in public shared environments according to preferences of the group of people present in there. In ambient intelligence contexts, we cannot assume that the system will be able to know every users physically present in the environment and therefore to access to their profiles in order to compute the preferences of the entire group. For this reason, we assume that group members may be i) totally unknown, ii) completely or iii) partially known by the system. As we describe in the paper, in the first case, the system uses a group profile that is built statistically according to the results of a preliminary study. In the second case, the model of the group is created from the profiles of known users. In the third situation the group interests are modeled by integrating preferences of known members with a statistical prediction of the interests of unknown ones. Evaluation results proved that adapting news display to the group was more effective in matching the members’ interests in all the three cases than the in the non-adaptive modality.

Berardina De Carolis, Sebastiano Pizzutilo

Backmatter

Additional information

Premium Partner

    Image Credits