1,720,957 research outputs found
Climaps by Emaps in 2 Pages (A Summary for Policy Makers and Busy People)
No full textClimaps.eu is an online atlas providing data, visualizations and commentaries about climate adaptation debate. It contains 33 issue-maps and 5 issue-stories. Each of the maps focuses on one issue in the adaptation debate and provides.
The atlas is addressed to climate experts (negotiators, NGOs and companies concerned by global warming, journalists...) and to citizens willing to engage with the issues of climate adaptation.
It employs advanced digital methods to deploy the complexity of the issues related to climate adaptation and information design to make this complexity legible
Weaving data, slicing views
No full textDigital archives metadata suggest a rich and complex system of relationships between the different properties of archived items, which is often not properly represented. Lomen is a research project aimed at exploiting the richness of digital archives, stitching up the relationships between entities and providing visual access to the system. This paper presents the design process used to create such visual access for architect Baldessari's historical archives. The research results in a digital platform that allows users to explore contents in a non-linear way, identifying patterns and fostering insight. The platform also aims at weaving together several levels of information through direct linking to archive entities such as projects, artifacts or individuals involved. Curators are also given the ability to elaborate theme-based paths, providing varied and unique entry points to the underlying data to users
Searching for dominant high-level features for music information retrieval
Full text linkMusic Information Retrieval systems are often based on the analysis of a large number of low-level audio features. When dealing with problems of musical genre description and visualization, however, it would be desirable to work with a very limited number of highly informative and discriminant macro-descriptors. In this paper we focus on a specific class of training-based descriptors, which are obtained as the loglikelihood of a Gaussian Mixture Model trained with short musical excerpts that selectively exhibit a certain semantic homogeneity. As these descriptors are critically dependent on the training sets, we approach the problem of how to automatically generate suitable training sets and optimize the associated macro-features in terms of discriminant power and informative impact. We then show the application of a set of three identified macro-features to genre visualization, tracking and classification
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Musical genre classification and tracking based on clustering driven high level features
Full text linkLAUREA SPECIALISTICAL'avvento dell'era digitale e la diffusione su larga scala di internet hanno reso i contenuti multimediali i principali veicoli di informazione e comunicazione di oggi. La fruizione di audio e video è cresciuta esponenzialmente, grazie alla facilità con cui sono creati, distribuiti e condivisi.
Parlando nello specifico dell'ambito musicale, sono stati sviluppati moltissimi servizi e software dedicati alla sua produzione, analisi, riproduzione e ricerca, dedicati a tutti i tipi di utenti (professionisti, amatori e utenti comuni). Il campo di ricerca che permette l'implementazione di queste tecnologie è chiamato Music Information Retrieval (MIR). Uno dei problemi più ostici di questa disciplina è il riconoscimento automatico del genere musicale, il cui obiettivo è quello di categorizzare automaticamente brani audio rispetto ad una tassonomia di classi.
La maggior parte dei lavori svolti in questa direzione usa descrittori di basso livello per descrivere un dataset di training e fornisce la distribuzione di punti di dati risultante ad un classificatore, che decide il genere musicale di dati senza etichetta attraverso un processo di apprendimento supervisionato. I descrittori di basso livello non hanno un significato semantico intellegibile, ma possono essere usati da sistemi informatici per la discriminazione tra classi. Ultimamente, tuttavia, parte della ricerca è mirata all'implementazione di sistemi che adottino caratteristiche di alto livello per descrivere un segnale audio; usando un livello di astrazione più elevato, i descrittori acquistano un significato musicale comprensibile anche all'utente finale, e possono aprire nuove possibilità di implementazione.
Un problema riguardante questo tipo di descirittori è la loro definizione, che è generalmente svolta in modo soggettivo, selezionando i descrittori di basso livello che intuitivamente dovrebbero definire una caratteristica musicale se combinati.
Questa tesi propone un metodo oggettivo e bassato su esempio per definire un insieme di feature di alto livello e le usa per la classificazione del genere musicale. La definizione dei descrittori di alto livello è ottenuta trovando il sottoinsieme di caratteristiche di basso livello che meglio clusterizza un dataset di training, e modellando i cluster risultanti attraverso un modello statistico usando un classificatore GMM. La fase di classificazione è invece svolta tramite un classificatore SVM, usando unicamente i descrittori di alto livello precedentemente implementati
Sono stati implemntati due approcci: il primo, seguendo un percorso bottom-up, è mirato a capire se cluster compatti e ben separati forniscano una classificazione più efficiente. Il secondo, seguendo un percorso top-down, è mirato a trovare il sottoinsieme di descrittori di basso livello che formi i cluster che portano alla accuracy più alta possibile.
I risultati ottenuti mostrano risultati promettenti, specialmente per quanto riguarda l'approccio top-down.The coming of the digital era and the large scale diffusion of internet made multimedia one of the main means of information and comunication of these days. The fruition of audio and video digital content has grown exponentially, thanks to their creation, distribution and sharing.
Talking specifically about music, a great quantity of services and software dedicated to its production, analysis, reproduction and research has been developed, dedicated to all kinds of users (professional, amateur and common ones). The field of research alllowing these technologies is called Music Information Retrieval (MIR). Among this interdisciplinary science, one of the most challenging tasks is the automatic musical genre classification; the aim of this problem is to automatically categorize audio excerpts according to a taxonomy of classes.
Usual works in this direction use low-level features to describe a training dataset and give the resulting data points distribution to a classifier, which classifies unlabeled data according to a supervised learning process. Such low-level descriptors have no immediate meaning to humans (experts or common users), but can be used by a machine to discriminate between classes. Lately, though, part of the research is focused on the implementation of systems adopting high-level features to describe audio signals; using a higher abstraction level, these descriptors acquire an intelligible music meaning; they could therefore be used by experts and common users, opening new implementation possibilities.
An issue in using these type of features is their definition, which is usually done in a subjective way by selecting the low-level descriptors which by guess should resemble a high-level characteristic if combined.
This thesis proposes an objective, example-based method to define a set of high-level features and use them to perform genre classification. The definition is obtained by finding the set of low-level features best clusterizing a training dataset, and modeling the resulting clusters through a statistical model using a GMM classifier. The classification step is instead performed with an SVM classifier, using only the derived high-level features.
Two approaches are implemented: one, following a bottom-up fashion, aims to find out if compact and well separated clusters deliver a more efficient classification. The other, following a top-down fashion, aims at finding the subset of low-level features forming the clusters that achieve the best classification.
Experimental results show promising results, especially for the top-down approach
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Full text linkWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
