Search

In many applications that rely on machine learning, the availability of labelled data is a matter of primary importance. However, when tackling new tasks, labels are usually missing and must be collected from scratch by the users. In this work, we address the problem of learning classifiers when the amount of labels is very scarce. We do so by learning multiple vectors, called prototypes, that represent relevant semantic concepts for the task at hand. We propose a theoretically inspired mechanism that computes probabilities of matching between the prototypes and the input elements, and we combine these probabilities to increase the expressiveness of the classifier. Moreover, by leveraging low-cost extra annotations in the training data, a simple error-boosting technique guides the learning process and provides substantial performance improvements. Empirical results confirm the benefits of the proposed approach in both balanced and unbalanced datasets. Our methodology is thus of practical use when gathering and labelling new examples is more expensive than annotating what we already have.

The adaptive processing of graph data is a long-standing research topic which has been lately consolidated as a theme of major interest in the deep learning community. The snap increase in the amount and breadth of related research has come at the price of little systematization of knowledge and attention to earlier literature. This work is designed as a tutorial introduction to the field of deep learning for graphs. It favours a consistent and progressive introduction of the main concepts and architectural aspects over an exposition of the most recent literature, for which the reader is referred to available surveys. The paper takes a top-down view to the problem, introducing a generalized formulation of graph representation learning based on a local and iterative approach to structured information processing. It introduces the basic building blocks that can be combined to design novel and effective neural models for graphs. The methodological exposition is complemented by a discussion of interesting research challenges and applications in the field.

In this paper we present ThReeNN, a model for Community Question Answering, Task 3, of SemEval-2017. The proposed model exploits both syntactic and semantic information to build a single and meaningful embedding space. Using a dependency parser in combination with word embeddings, the model creates sequences of inputs for a Recurrent Neural Network, which are then used for the ranking purposes of the Task. The score obtained on the official test data shows promising results.

We propose a new Graph Neural Network that combines recent advancements in the field. We give theoretical contributions by proving that the model is strictly more general than the Graph Isomorphism Network and the Gated Graph Neural Network, as it can approximate the same functions and deal with arbitrary edge values. Then, we show how a single node information can flow through the graph unchanged.

The awareness of the visitor’s role centrality in museums has led to the constant development of techniques and methodologies aimed at improving the visitor’s experience. In this context, Augello et al. suggest the four pillars to enhance the user experience: User Localization, User Understanding, Multimodal Interaction, and Gamification. As part of our research activities, we have focused on the latter two pillars. More specifically, User Localization allows museum curators and staff to understand the user movements in the museum and how much time she spends in front of the artworks. This information can help personalize the user experience to make it more and more satisfying. To collect this information, we believe that rather than using traditional techniques, such as beacons or IR, it is better to leverage deep learning techniques using cameras, thus supporting subsequent analyzes such as an automatic analysis of the user’s face. In particular, analyzing the user’s face allows us to grasp her reactions in front of the artworks and establish whether she may have liked that artwork or not, and if she may want to see similar artworks. The aim of this research is to understand how to use localization systems and automatic face analysis to build a user model that could allow museum staff to customize her experience. The first results obtained regarding an innovative localization system, based on deep learning, show the potential of this technique, which permits the extraction of information from faces through the same cameras. As regards this second part, techniques based on deep neural networks, such as Convolutional and Recurrent Neural Networks, are being tested to identify the best approach to extract implicit feedback from users’ faces without making them fill out long questionnaires that require active participation and steal time from the visit.