Data Management in Pervasive Systems

The new capacities of pervasive and ubiquitous computing are defining new horizons for human creativity and connectivity. This is also because pervasive and ubiquitous computing is very often linked with other emerging technologies, such as the semantic Web, cloud computing, and affective computing. At the same time, there has been a rapid development in 3G and 4G networks as well as in the use of feature-rich smartphones. This means that everyone can be connected at anytime and anywhere. In the near future, the development of specific hardware as well as software will enable everyone to be in touch with everything and everywhere, thus closing the “circle of pervasiveness.” The Internet of the future promises to connect our mobile devices with everything (from the fridge in our homes to special sensors in our cars or even in our bodies), whenever and whereverwe are.

In today’s society, most actions we perform are recorded and the collected data are stored, processed, and possibly shared in a way that was impossible until a few years ago before the development of ubiquitous and pervasive technologies. Ubiquitous technologies represent one of the most significant revolutions in information and communication technologies.

Wireless sensor networks (WSNs) are a clear example of the disappearing technology expressed by Mark Weiser [49] since they allow humans to interact with the environment, feeding the user with a large set of heterogeneous data. At the same time, the complexity of the envisaged WSN-based systems grew from a handful of homogeneous sensors to hundreds or thousands of devices differing as to their capabilities, technologies, architectures, and languages.

The massive usage of data streams dates back to artificial satellite information processing systems and to their commercial application in the early 1970s, such as in telecommunications switching, land monitoring, meteorological surveillance, etc. Today they are extensively used in monitoring systems applications based on wired and wireless sensor networks, in social networks, and in the Internet of Things [20]. The main functional goals of data stream management systems (DSMSs) are as follows: (a) results must be pushed to the output promptly and eagerly while input tuples continue to arrive and (b) because of the unbounded and massive nature of data streams, all past tuples cannot be memorized for future use. Only synopses can be kept in memory and the rest must be discarded.

Complex event processing (CEP) systems represent a mainstream approach for processing streams of data. Specifically, they target the definition and detection of high-level situations of interest, or composite events, starting from streams of primitive events collected from the external environment. In CEP, composite events are specified through user-defined rules, which express how to select, manipulate, and combine primitive events. Thanks to the capability of handling large volumes of information to isolate situations of interest, CEP represents a perfect solution for the management and online analysis of data in pervasive systems. Researchers and practitioners working on CEP focused on the creation of simple, yet expressive languages for the definition of CEP rules. At the same time, they also put significant effort on performance and scalability, defining efficient algorithms and rule evaluation mechanisms that enable high-throughput and low processing delay. This chapter focuses on both aspects. On the one hand, it presents the processing abstractions offered by existing CEP systems in details, focusing on the applicability to pervasive systems. On the other hand, it presents some of the processing algorithms and techniques that contribute to the performance of state-of-the-art CEP systems.

More and more applications require real-time fine-grained query answering on massive, heterogeneous, noisy and incomplete data streams. Indeed, systems capable of scalable stream processing exist. Specialised data stream management systems (DSMSs) and complex event processing (CEP) have been largely investigated in the 2000s. They can provide reactive fine-grained information access even in the presence of noisy data. What they lack is the ability to master heterogeneity and incompleteness. In this chapter, we show out to apply semantic interoperability principles to data streams. In particular, we described recently developed methods that use extensions of semantic Web technologies (i.e. RDF, SPARQL and OWL) to continuously answer fine-grained query on heterogeneous and incomplete data streams in a scalable manner. To make the chapter easier to follow, examples are provided in the context of sensor network and social media analytics.

In this chapter, we describe the main aspects of multimedia social networks (MSNs). We then describe the project global interactive virtual archaeological system (GIVAS), a multimedia environment for archaeologists, cultural heritage researchers, and normal tourists that provides a comprehensive collaborative platform for managing, searching, visualizing, and sharing multimedia cultural heritage information, using an innovative social network approach.

The spread of social networks as Twitter, Facebook, or Google+ or specialized ones as LinkedIn or Viadeo allows sharing opinions on different aspects of life every day. Millions of messages appear daily on the web thanks to blogs, microblogs, social networks, or review collector sites. This textual information can be divided in two main categories: facts and opinions. Facts are objective statements, while opinions reject and reveal people’s sentiments about products, personalities, and events and are extremely important when someone needs opinions before taking a decision. This information is a rich source of data for opinion mining. The interest that potential customers show in online opinions and reviews about products is something that vendors are gradually paying more and more attention to. In this scenario, a promising approach is sentiment analysis: the computational study of opinions, sentiments, and emotions expressed in a text. Its main aim is the identification of the agreement or disagreement statements that deal with positive or negative feelings in comments or reviews. In this chapter, we investigate the literature’s state of the art and the adoption of a probabilistic approach based on the latent Dirichlet allocation (LDA). By this approach, for a set of documents belonging to a same knowledge domain, a graph, the mixed graph of terms (mGTs), can be automatically extracted. This graph contains a set of weighted word pairs, which are discriminative for sentiment classification. The proposed method has been applied for the real-time analysis of documents, as TripAdvisor’s posts, in Italian language of opinion holders or social groups in the Databenc context: urban spaces, museums, archaeological parks, and events.

Online social networks (OSNs) have become extremely popular in recent years, and their widespread adoption has led to the presence of huge volumes of users’ personal information on the Internet. The ever-increasing number of social networks’ users on one hand and the massive amount of information being shared daily on the other hand have encouraged attackers to develop and use different techniques to collect and analyze such information for a number of malicious purposes, including spear-phishing attacks and identity theft. Clearly, this trend represents a significant challenge for both users and administrators. In fact, the widespread adoption of OSNs has raised a wide range of security and privacy concerns, which have not been fully addressed yet. In many cases, users are not even aware of the disclosure of their personal information through their profiles. Leakage of a user’s private information can happen in different ways. In this chapter, we discuss the main security and privacy issues associated with online social networks and investigate some attack models used to reveal a user’s private information. We also discuss different strategies and regulations that can prevent disclosure of private information through OSNs.

The massive data generated by devices and by people creates an enormous number of possibilities for applications and services. As the combinations of data, services and applications explodes, personalization comes as a meaningful way to filter data and present it in useful ways to users taking into account both the individual features of each user or characteristics of groups of users as well as the desired properties of the selected data, such as serendipity, coverage and diversity. In this chapter, we will start by providing a broad definition of personalization. We will examine different forms of data personalization, ranging from customization to recommendations to personalized search and exploration. We will describe properties of data personalization, such as serendipity and diversity, and we will present representative methods that achieve personalized results of different properties. Then, we will describe how preferences, either at the group or the individual level can be modelled, and we will conclude the chapter with an overview of user preference learning methods.

Context represents any information that can be used to characterize the situation of an entity. An entity is a person, place, or object that is considered relevant to the interaction between a user and an application, including the user and application themselves. The ubiquity of mobile devices (e.g., smartphones, GPS devices) has in part motivated the use of contextual information in modern mobile applications. From one perspective, context in mobile systems can fall into three categories: (a) user context that includes the personal attributes of the user, e.g., spatial location and budget; (b) point-of-interest (POI) context, e.g., restaurant location, operating time, and rating; and (c) environmental context, e.g., weather and road conditions. Incorporating such context in applications provided to mobile users may significantly enhance the quality of service in terms of finding more related answers. This chapter first gives a brief overview of context and context awareness in mobile systems. It then discusses different ways of expressing the spatial location context within mobile services. The chapter later describes three main application examples that can take advantage of various mobile contexts, namely, social news feed, microblogging (e.g., Twitter) and recommendation services. The chapter finally presents a generic method that incorporates context and user preference awareness in database systems—which may serve as a backbone for context-aware mobile applications.

Content-based multimedia information retrieval (IR) provides new models and methods for effectively and efficiently “searching” through the huge variety of media that are available in different kinds of repositories (digital libraries, Web portals, social networks, multimedia databases, etc.). In this chapter, we will review the current state of the art of content-based multimedia information retrieval, including the most promising browsing and search paradigms for the several types of multimedia data, and show some cultural heritage applications.

The intrinsic complexity and diversity of data in multimedia digital libraries (MDLs) require devising techniques and solutions that are inherently different from those usually adopted in traditional information retrieval and database (DB) systems. Moreover, the size and the dynamicity of MDLs force researchers to strive for efficiency, so as to guarantee real-time results to the users. Finally, semantics should be also brought into context, in order to facilitate users’ experience in querying, browsing, and consuming multimedia information. This chapter will present an approach toward the efficient, effective, and semantically rich data retrieval in MDLs. With respect to the commonly used holistic approach, where the multimedia datum is considered as an atomic entity, our reductionist strategy considers the multimedia information as a complex combination of component subparts and eases the fulfillment of the three above properties of efficiency, effectiveness, and semantic richness. Indeed, by decomposing multimedia information into simpler and smaller component objects, we are able to index such components without giving up the ability of querying the original information as a whole.

In the last decade, the spread of broadband Internet connections even for mobile devices has contributed to an increased availability of multimedia information on the Web. At the same time, due to the decrease of storage cost and the increasing popularity of storage services in the cloud, the problem of information overload has become extremely serious even in personal/company archives. The need to manage, retrieve, and present all these data has promoted the development of advanced multimedia information systems, which include recommendation modules to account for the requests of personalized data selection and presentation. Recommendation systems estimate ratings, or utilities, which quantify users’ degree of interest for the different available data, so that the data can be offered to the users in a personalized way, in decreasing order of interest. Multiple approaches have been proposed in the literature to estimate such degrees of interest. In content-based filtering, the utility (for a user) of a given item is estimated as a function of the ratings given by the same user to other similar items. A dual approach is collaborative filtering, in which filtering (i.e., estimating the object’s utilities) for a given user is computed by referring to the opinions of other similar users. Thus, a major challenge faced by collaborative filtering is the need to associate each user to a set of other users having similar profiles. In this chapter, we first present the co-clustering-based recommendation techniques, which allow to combine heterogeneous multimedia content information and data about the users’ preferences and rankings, thus overcoming some of the content-based filtering drawbacks, as well as some collaborative filtering weaknesses. Then, we briefly discuss the challenges in multimedia delivery and the most common strategies adopted in the context of cultural heritage media delivery.

In this final chapter, we show an application of discussed data management pervasive technologies for a cultural heritage smart scenario by describing the PATCH (Portable context-aware ATlas for Cultural Heritage) system—a “portable” prototype of a multimedia and social “atlas” of cultural points of interest, in which browsing is driven by the context. The system is characterized by several features typical of pervasive systems: (1) capability of gathering information from distributed and heterogeneous pervasive data sources (e.g., sensor networks, social networks, digital libraries and archives, multimedia collections, web data services, etc.); (2) context awareness and, consequently, ability of providing useful and personalized data and services for users on the basis of their preferences and of the surrounded environment; (3) advanced data management techniques and technologies necessary to deal with the information variety, velocity, and volume; and (4) advanced smart services as retrieval, recommendation, analytics, and other utilities. A specific application of our system within DATABENC is proposed by means of a real case study related to the Paestum Ruins.