Archivio Istituzionale della Ricerca- Università degli Studi di Foggia
Not a member yet
55976 research outputs found
Sort by
Organizzazione mondiale del commercio e intelligenza artificiale
No full textAbstract. — Il contributo analizza in prospettiva sistematica il rapporto tra l’Organiz-
zazione mondiale del commercio e l’intelligenza artificiale, mettendo in luce come l’IA operi
oggi come cross-cutting issue nell’ambito del sistema commerciale multilaterale. In assenza di
norme specificamente dedicate, il sistema OMC disciplina l’IA mediante una costellazione di
disposizioni contenute in GATT, GATS, TRIPs, TBT e negli strumenti sul commercio
elettronico, rilevanti lungo l’intero ciclo di vita dell’IA. L’articolo esamina criticamente i
problemi di qualificazione dei prodotti smart e dei servizi di IA, le incertezze interpretative
legate alla distinzione beni/servizi, la rilevanza degli ostacoli tecnici agli scambi, nonché
l’impatto delle clausole di eccezione — in particolare quelle di sicurezza nazionale — sempre
più invocate nei contesti digitali. Sono inoltre analizzati gli aspetti di proprietà intellettuale, nel
contesto TRIPs, con attenzione all’uso di dataset protetti e alle difficoltà di inquadrare
giuridicamente gli output generati da IA. L’analisi mostra come il principio di neutralità
tecnologica consenta di adattare il sistema esistente, ma non possa sostituire indefinitamente
norme specifiche; si suggerisce infine che un modello progressivo e plurilaterale, simile a
quello che ha condotto all’AEC, possa rappresentare una via percorribile per una futura
regolazione internazionale dei profili commerciali dell’IAAbstract. — This paper provides a systematic assessment of the relationship between the
World Trade Organization and artificial intelligence, showing how AI has become a “cross‐cut-
ting” issue within international trade law. In the absence of AI-specific rules, the WTO
framework governs the technology through a dispersed set of provisions in the GATT, GATS,
TRIPs, TBT Agreement, and electronic commerce instruments, all of which are relevant to
different stages of the AI lifecycle. The article examines the legal challenges posed by the
qualification of smart products and “AI‐driven” services, the persistent uncertainty created by the
goods/services dichotomy, and the growing relevance of technical barriers to trade, as well as the
increasing recourse to general and national security exceptions in digital contexts. Intellectual
property issues are also explored in TRIPs context, including the use of protected datasets for
training and the complex question of copyrightability of “AI‐generated” outputs. The analysis
concludes that, while technological neutrality principle enables an evolutionary interpretation of
existing WTO rules, it cannot indefinitely compensate for the lack of AI-specific provisions. A
gradual, plurilateral approach-modelled on the process that led to the 2024 Agreement on
Electronic Commerce-may offer a viable path toward future international regulation of “AI‐re-
lated” trade matter
Stabilizing Inference in Dirichlet Regression via Ridge-Penalized Model
No full textWe propose a penalized Dirichlet regression framework for modeling compositional data, using a softmax link to ensure that the mean vector lies on the simplex and to avoid log-ratio transformations or zero replacement. The model is formulated in a GLM-like setting and incorporates an mathematical equation (ridge) penalty on the regression coefficients to improve stability in the presence of multicollinearity, high-dimensional covariate spaces and weak effects. The classical Dirichlet regression is recovered as the special case with zero penalty, so that the proposed estimator nests the standard approach. Estimation is carried out via a gradient-based block coordinate ascent algorithm, for which we derive closed-form expressions for the log-likelihood gradient and for the Jacobian of the softmax transformation. We investigate the performance of the method through a simulation study that includes orthogonal and correlated designs, different noise levels and a sparse high-dimensional scenario. The results show that the penalized and unpenalized estimators are essentially equivalent in simple well-posed settings, while the ridge-penalized model achieves systematically lower coefficient RMSE and higher log-likelihood in more challenging configurations. Finally, we apply the method to US male cause-of-death data, where cross-validated penalization yields improved fit and smooth, interpretable age–cause profiles supported by bootstrap confidence intervals
Adesione a PVC e forme di tutela
No full textIl contributo esamina la rinnovata disciplina dell’adesione ai processi verbali di constatazione, introdotta dall’art. 5-quater del D.Lgs. n. 218/1997, soffermandosi sui relativi riflessi in tema di accesso anticipato alla giurisdizione tributaria. Muovendo dalla progressiva erosione della tipizzazione degli atti impugnabili di cui all’art. 19 del D.Lgs. n. 546/1992, si intende dimostrare la possibilità di impugnare il p.v.c. senza necessità di attendere l’emanazione dell’avviso di accertamento, in particolare ove emergano errori manifesti non emendati. In tale prospettiva, si evidenzia come il verbale presenti un’attitudine immediata a incidere sulla sfera giuridica del contribuente, facendo sorgere un interesse ad agire concreto e attuale, tale da rendere non procrastinabile l’accesso all’Autorità giudiziaria
Intro - L'automatizzazione della discriminazione: una prospettiva giuridica multidisciplinare
No full textRising from the pandemic: do ESG factors matter for resiliency in the Italian Stock Market?
No full textThis study examines the resilience of ESG investments across
different phases of the COVID-19 pandemic, focusing on the Italian Stock
Exchange. We compare the performance of firms included in the MIB ESG
index with a carefully stratified control group of non-ESG companies. Using a
network analysis approach, the results show a higher degree of resilience
among ESG assets, particularly in the post-lockdown phase, when they
exhibited a faster and stronger recovery than traditional investments. The
findings suggest that firm-level commitments to environmental responsibility,
social engagement, and sound governance can enhance market robustness
during periods of uncertainty and stress. By providing empirical evidence on
the stabilising role of ESG practices, the study contributes to the sustainable
finance literature and improves understanding of the relationship between ESG
investing and market dynamics. The analysis identifies consumer discretionary,
healthcare, and technology as the sectors that benefited most from ESG
adoption during the recovery phase
Selection and characterization of salt-tolerant plant growth promoting bacteria associated with the endosphere and rhizosphere of perennial glasswort from the Apulia Region (Italy)
No full textUncovering network changes in the evolution of an innovation niche
No full textThe present paper aims at unravelling the networking dynamics regarding the evolution of an innovation-niche, by looking at its network configuration over time. To this end, we investigate the actors’ role within the network as well as the network architecture. We employ the social network analysis (SNA) to four different but interrelated types of networks characterising the Italian biofuel industry. Each network was investigated in both its development and maturity phase. Our findings show that, over the niche evolution, actors established new relations, increasing networks’ density and making the networks more centralized. Furthermore, while two out of four networks confirmed their small world configuration over time, one (i.e. “communication” network) increased its small-world-ness, providing a larger number of information channels and more social reinforcement chances for niche actors to innovate. Overall, results suggest that policy makers should accompany and ease the innovation-niche evolution path by fostering the participation of the laggard behind and actors' clusterization, moderating any possible lock-in risk
Effect of coal ash fly zeolite and Bacillus subtilis on water-use efficiency of chickpea grown under water deficit
No full textChickpea cultivation may be considered strategic for improving sustainability in Mediterranean cropping systems; however, global warming and drought may pose limitations to yield stability. Soil amendments, such as zeolite (Z), and microbial inoculation with plant growth-promoting bacteria (PGPB) are indicated as potential strategies to mitigate water deficiency. Still, their combined application has been little investigated, especially in pulse crops. To evaluate the effects of synthetic zeolites, PGPB inoculation, and water deficit on chickpea, a pot experiment was carried out under controlled conditions. Two chickpea genotypes (Pascià and Sultano) were subjected to a factorial combination of 2% soil-added zeolite and PGPB under optimal and limiting water conditions (100 and 50% of the water requirement, respectively). The results showed that water supply strongly influenced crop response, with a reduction of about 50% in grain yield under water deficit. Treatments with zeolite and PGPB (Bacillus subtilis) inoculation led to a significant increase in yield and water-use efficiency (WUE) on both chickpea genotypes under water deficit, and their combination further improved harvest index. Under non-limiting water conditions, the increase in WUE was significant with PGPB inoculation. An increase in protein content (PC) was also observed in plants treated with zeolite under water-deficit conditions, while grain total phenolic content (TPC) was negatively influenced by water supply. Overall, agronomic treatments mitigated the effects of water deficit: zeolite improved harvest index and nitrogen uptake while PGPB enhanced WUE. Their combined use appears promising as a strategy to alleviate the negative impacts of water deficit on chickpea cultivation