Marco Grassi

Sentic Web: A New Paradigm for Managing Social Media Affective Information

The recent success of media-sharing services caused an exponential growth of community-contributed multimedia data on the Web and hence a consistent shift of the flow of information from traditional communication channels to social media ones. Retrieving relevant information from this kind of data is getting more and more difficult, not only for their volume, but also for the different nature and formats of their contents. In this work, we introduce Sentic Web, a new paradigm for the management of social media affective information, which exploits AI and Semantic Web techniques to extract, encode, and represent opinions and sentiments over the Web. In particular, the computational layer consists in an intelligent engine for the inference of emotions from text, the representation layer is developed on the base of specific domain ontologies, and the application layer is based on the faceted browsing paradigm to make contents available as an interconnected knowledge base.

Sentic Computing for social media marketing

In a world in which millions of people express their opinions about commercial products in blogs, wikis, fora, chats and social networks, the distillation of knowledge from this huge amount of unstructured information can be a key factor for marketers who want to create an image or identity in the minds of their customers for their product, brand or organization. Opinion mining for product positioning, in fact, is getting a more and more popular research field but the extraction of useful information from social media is not a simple task. In this work we merge AI and Semantic Web techniques to extract, encode and represent this unstructured information. In particular, we use Sentic Computing, a multi-disciplinary approach to opinion mining and sentiment analysis, to semantically and affectively analyze text and encode results in a semantic aware format according to different web ontologies. Eventually we represent this information as an interconnected knowledge base which is browsable through a multi-faceted classification website.

Developing HEO human emotions ontology

A big issue in the task of annotating multimedia data about dialogs and associated gesture and emotional state is due to the great variety of intrinsically heterogeneous metadata and to the impossibility of a standardization of the used descriptor in particular for the emotional state of the subject. We propose to tackle this problem using the instruments and the vision offered by Semantic Web through the development of an ontology for human emotions that could be used in the annotation of emotion in multimedia data, supplying a structure that could grant at the same time flexibility and interoperability, allowing an effective sharing of the encoded annotations between different users.

Towards an ontology framework for intelligent smart home management and energy saving

Energetic efficiency has become a mandatory requirement for buildings. Several efforts have been done both to reduce energy consumption and to promote alternative generation sources. In most of the existing proposals these two facets of energy conservation are handled singularly. We believe that energy production and consumption need to be managed in a unique perspective to enable a more efficient energy administration. In this paper, we briefly introduce a novel holistic vision for smart home environments and present an ontology framework conceived to provide the necessary information modelling in its implementation. In particular, we focus on two of the composing ontologies that we are currently developing for describing devices and power generation/consumption, in order to enable intelligent task management for energy consumption optimisation.

Pundit: augmenting web contents with semantics

Scholars are using the Web every day to search, read, collaborate, and ultimately do their research. While some of the basic activities that the scholars do, such as reading and writing papers, are already well supported in the digital world, some essential scholarly primitives, such as annotation, augmentation, contextualiza- tion, and externalization, do not yet have clear support in terms of software tools. What scholars ultimately do during their research activity is to iteratively and collaboratively create new knowledge. With the advent of the Digital Humanities, we now have the opportunity—and technology—to capture at least a part of this knowledge and make it available as machine-processable data so to be better explorable and discoverable. In this paper, we present and discuss Pundit: a novel semantic annotation tool that enables scholars to collect, annotate, and contextualize Web resources. Deep-linking is used in conjunction with an RDF- based data model to allow granular selection of content (e.g. text excerpts, image fragments). Pundit aims at enabling scholars to produce meaningful machine- readable data that captures the semantics of their annotations. By providing a customizable annotation environment, where domain specific vocabularies can be loaded, and easy ways of integrating with existing Web archives or libraries, Pundit enables users to publish their annotations and collaboratively build a semantic graph. Such a graph can be consumed via HTTP APIs and standard SPARQL, thus allowing existing Linked Data applications to easily work with the data and Web clients in general to build specific visualizations.

Ontologies for Smart Homes and energy management: An implementation-driven survey

Semantic Web technologies have become a reference technology for information modelling and reasoning support in Smart Homes. This paper provides an extensive review of the ontologies developed in this scenario. Also, it discusses how they can be connected and expanded to create a complete framework that covers all the aspects of a Smart Home, ranging from device description to energy management, under a unifying holistic vision.

A Collaborative Video Annotation System Based on Semantic Web Technologies

In recent years, videos have become more and more a familiar multimedia format for common users. In particular, the advent of Web 2.0 and the spreading of video-sharing services over the Web have led to an explosion of online video content. The capability to provide broader support in accessing and exploring video content, and in general other kind of multimedia formats as images and documents, is becoming more and more important. In this context, the value of semantically structured data and metadata is recognized as a key factor both to improve search efficiency and to guarantee data interoperability. This latter aspect is critical to connect different, heterogeneous content coming from a variety of data sources. On the other hand, the annotation of video resources has been increasingly understood as a medium factor to enable deep analysis of contents and collaborative study of online digital objects. However, as existing annotation tools provide poor support for semantically structured content or in some cases express the semantics in proprietary and non-interoperable formats, such knowledge that users build by carefully annotating contents hardly crosses the boundaries of a single system and often cannot be reused by different communities (e.g., to classify content or to discover new relations among resources). In this paper, a novel Semantic Web-based annotation system is presented that enables user annotations to form semantically structured knowledge at different levels of granularity and complexity. Annotation can be reused by external applications and mixed with Web of Data sources to enable “serendipity,” the reuse of data produced for a specific task (annotation) by different people and in different contexts from the one data originated from. The main ideas behind the approach are to build on ontologies and support linking, at data level, to precise thesauri and vocabularies, as well as to the Linked Open Data cloud. By describing the software model, developed in the context of SemLib EU project, and by providing an implementation of an online video annotation tool, the main aim of this paper is to demonstrate how such technologies can enable a scenario where users annotations are created while browsing the Web, naturally shared among users, stored in machine readable format and then possibly recombined with external data and ontologies to enhance end-user experience.

Towards an RDF encoding of ConceptNet

The possibility of relying on a rich background knowledge constitutes a key element for the development of more effective sentiment analysis and SW applications. In this paper, we propose to encode the wide knowledge base collected by Open Mind Common Sense initiative into ConceptNet, in a semantic aware format to make it directly available for Semantic Web applications. We also discuss how the encoding of ConceptNet into RDF can be beneficial to promote its connection with other resources such as WordNet and HEO ontology to further extend its knowledge base.

Towards a semantically-enabled holistic vision for energy optimisation in smart home environments

Nowadays energy-saving represents a mandatory requirement for building. Several efforts have been done both to reduce energy consumption and to promote alternative generation sources. By the way, in most of the existing systems these two faces of energy-conservation are managed in isolation. In this paper, we propose a novel holistic vision for a smart home environment, in which energy administration embraces both energy production and consumption and its handled in conjunction with services management. The proposed system uses an IP-based network as main communication channel and a semantic extension of the UPnP protocol for devices auto-configuration and control. An ontology framework is used to encode all the relevant information about devices, services and context, including energy status and user preferences, into a global knowledge base. Using the inference capabilities provided by the used rich semantic descriptions, such gKB can be used to support efficient control logics and intelligent decision making, that can be exploited also for a more effective energy management.

Gender recognition using PCA and DCT of face images

In this paper we propose a gender recognition algorithm of face images. We have used PCA and DCT for dimensionality reduction. The algorithm is based on a genetic algorithm to improve the selection of training set of images for the PCA algorithm. Genetic algorithm helps to select the images, which best represent each gender, from the image database. We have evaluated a nearest neighbor classifier as well as a neural network. Experimental results show a correct identification rate of 85,9%.