Entity

Time filter

Source Type

Trento, Italy

Balahur A.,European Commission - Joint Research Center Ispra | Turchi M.,Fondazione Bruno Kessler
Computer Speech and Language | Year: 2014

Sentiment analysis is the natural language processing task dealing with sentiment detection and classification from texts. In recent years, due to the growth in the quantity and fast spreading of user-generated contents online and the impact such information has on events, people and companies worldwide, this task has been approached in an important body of research in the field. Despite different methods having been proposed for distinct types of text, the research community has concentrated less on developing methods for languages other than English. In the above-mentioned context, the present work studies the possibility to employ machine translation systems and supervised methods to build models able to detect and classify sentiment in languages for which less/no resources are available for this task when compared to English, stressing upon the impact of translation quality on the sentiment classification performance. Our extensive evaluation scenarios show that machine translation systems are approaching a good level of maturity and that they can, in combination to appropriate machine learning algorithms and carefully chosen features, be used to build sentiment analysis systems that can obtain comparable performances to the one obtained for English. © 2013 Elsevier Ltd. Source


Jurman G.,Fondazione Bruno Kessler
PloS one | Year: 2012

The outcome of a functional genomics pipeline is usually a partial list of genomic features, ranked by their relevance in modelling biological phenotype in terms of a classification or regression model. Due to resampling protocols or to a meta-analysis comparison, it is often the case that sets of alternative feature lists (possibly of different lengths) are obtained, instead of just one list. Here we introduce a method, based on permutations, for studying the variability between lists ("list stability") in the case of lists of unequal length. We provide algorithms evaluating stability for lists embedded in the full feature set or just limited to the features occurring in the partial lists. The method is demonstrated by finding and comparing gene profiles on a large prostate cancer dataset, consisting of two cohorts of patients from different countries, for a total of 455 samples. Source


Binosi D.,Fondazione Bruno Kessler | Chang L.,University of Adelaide | Papavassiliou J.,University of Valencia | Roberts C.D.,Argonne National Laboratory
Physics Letters, Section B: Nuclear, Elementary Particle and High-Energy Physics | Year: 2015

Within contemporary hadron physics there are two common methods for determining the momentum-dependence of the interaction between quarks: the top-down approach, which works toward an ab initio computation of the interaction via direct analysis of the gauge-sector gap equations; and the bottom-up scheme, which aims to infer the interaction by fitting data within a well-defined truncation of those equations in the matter sector that are relevant to bound-state properties. We unite these two approaches by demonstrating that the renormalisation-group-invariant running-interaction predicted by contemporary analyses of QCD's gauge sector coincides with that required in order to describe ground-state hadron observables using a nonperturbative truncation of QCD's Dyson-Schwinger equations in the matter sector. This bridges a gap that had lain between nonperturbative continuum-QCD and the ab initio prediction of bound-state properties. © 2015. Source


Patent
Airbus, Fondazione Bruno Kessler and Alenia Aermacchi | Date: 2012-09-21

A computer-implemented method for determining a configuration of a plurality of components in a systems installation which satisfies one or more constraints.


Grant
Agency: Cordis | Branch: H2020 | Program: RIA | Phase: EURO-6-2015 | Award Amount: 3.63M | Year: 2016

A seamless interaction with the public administration (PA) is crucial to make the daily activities of companies and citizens more effective and efficient, saving time and money in the management of administrative processes. In particular, online public services have an enormous potential for reducing the administrative burden of companies and citizens, as well as for creating saving opportunities for the PA. This potential is however far from being fully exploited. Online services made available by the PA typically rely on standardized processes, copied from their offline counterparts and designed only from the public sector organizations own perspective. This results in online services that fail to adapt to the specific needs of citizens and companies. With SIMPATICO, we address the issues above by proposing a novel approach for the delivery of personalized online services that, combining emerging technologies for language processing and machine learning with the wisdom of the crowd, makes interactions with the PA easier, more efficient and more effective. SIMPATICO combines top-down knowledge of the PA with bottom-up contributions coming from the community. These contributions can be of different types, ranging from the qualified expertise of civil servants and professionals to problems and doubts raised by citizens and companies that find online services difficult to use. Our approach is able to take into account both explicit information sources coming from citizens, professionals and civil servants, and implicit ones, extracted from user logs and past user interactions. SIMPATICOs learning by doing approach will use this information and match it with user profiles to continuously adapt and improve interactions with the public services. All the collected information on public services and procedures will be made available within Citizenpedia, a collective knowledge database released as a new public domain resource.

Discover hidden collaborations