1,721,038 research outputs found
Codici sematici e variabilità psicofisiologica in soggetti con attacchi di panico
No full texti soggetti con attacchi di panico presentano al livello psicofisiologico particolari network di attivazione del sistema vegetativo e di elaborazione corticale rispetto a stimoli emotigeni e a livello cognitivo, modalità espressive correlate all'alessitimia ed a una difficoltà di elaborazione ed espressione dei contenuti emotivi. obiettivo dello studio è stato quello di analizzare parte dei network vegetativi e corticali durante la rievocazione e narrazione di un evento di routine (ER)e di un evento rappresentativo dell'emozione paura (EP) in soggetti con DAP e di analizzare i codici semantici correlati alle narrazioni stesse. si sono analizzati i seguenti indici: GSR K (tonico) e GSR F (fasico) e HRV (frequenza cardiaca). rispetto agli indici cognitivi si è analizzato la durata del tempo di narrazione (TN) ei i lemmi associati alle descrizioni degli eventi negativi e positivi.
Risultati: i soggetti cin DAP presentano un range inferiore di variabilità per il GSR K nelle condizioni ER e EP; risulta invece tendente alla significatività la HRV nella condizione ER. Il TN è minore (significativamente) nella condizione EP . si potrebbe attribuire il minor tempo di narrazione per EP a una componente cognitiva di evitamento della narrazione con una valenza emozionale. Rispetto ai codici semantici il gruppo DAP non ha utilizzato lemmi emotigeni nel compito di rievocazione e descrizione di un evento rappresentativo dell'EP
Contextualized BERT Sentence Embeddings for Author Profiling: The Cost of Performances
No full textThe necessity to know information about the real identity of an online subject is a highly relevant issue in User Profiling, especially for analysis from digital sources such as social media. The digital identity of a user does not always present explicit data about her offline life such as age, gender, work, and more. This problem makes the task of user profiling complex and incomplete. For many years this issue has received a considerable amount of attention from the whole community, which has developed several solutions, also based on machine learning, to estimate user characteristics. The increasing diffusion of deep learning approaches has allowed, on the one hand, to obtain a considerable increase in predictive performance, but on the other hand, to have available models that cannot be interpreted and that require very high computational power. Considering the validity of new pre-trained language models on extensive data for resolving many natural language processing and classification tasks, we decided to propose a BERT-based approach (BERT-DNN) also for the author profiling task. In a first analysis, we compared the results obtained by our model with them of more classical approaches. As a follow, a critical analysis was carried out. We analyze the advantages and disadvantages of these approaches also in terms of resources needed to run them. The results obtained by our model are encouraging in terms of reliability but very disappointing if we consider the computational power required for running it
Simultaneous characterization of bulk impurities and interface states by photocurrent measurements
No full textA new method for evaluating both surface recombination velocity and bulk minority carrier lifetime by photocurrent measurements is discussed and validated by comparison with capacitance–voltage measurements of interface state density. This method is an evolution of the measurement of surface recombination velocity by the Elymat technique, it does not require the oxide to be etched off and consists in measurements of surface recombination velocity under an applied surface bias. The
application of a surface bias allows the control of the interface potential and the identification of the suitable interface condition
so that surface recombination velocity can be considered as a measurement of interface state density. In addition, it is shown that surface recombination velocity is suppressed when the surface is under accumulation conditions, so the application of a surface bias provides the possibility of a surface passivation by driving the surface into accumulation. This passivation by surface polarization is about as effective as the chemical passivation by HF. Finally, the dependence of surface recombination velocity on the injection level is shown to be reversed when the interface changes from depletion to accumulation or inversion conditions. This technique does not require the formation of a capacitor structure, so it is suitable for the measurement of as-grown interface properties. For this reason, this technique was chosen for a systematic study of the nitridation process of oxide films. Surface recombination velocity was correlated with nitrogen concentration at the oxide–silicon interface
Time of your hate: The challenge of time in hate speech detection on social media
Full text linkThe availability of large annotated corpora from social media and the development of powerful classification approaches have contributed in an unprecedented way to tackle the challenge of monitoring users' opinions and sentiments in online social platforms across time. Such linguistic data are strongly affected by events and topic discourse, and this aspect is crucial when detecting phenomena such as hate speech, especially from a diachronic perspective. We address this challenge by focusing on a real case study: the "Contro l'odio" platform for monitoring hate speech against immigrants in the Italian Twittersphere. We explored the temporal robustness of a BERT model for Italian (AlBERTo), the current benchmark on non-diachronic detection settings. We tested different training strategies to evaluate how the classification performance is affected by adding more data temporally distant from the test set and hence potentially different in terms of topic and language use. Our analysis points out the limits that a supervised classification model encounters on data that are heavily influenced by events. Our results show how AlBERTo is highly sensitive to the temporal distance of the fine-tuning set. However, with an adequate time window, the performance increases, while requiring less annotated data than a traditional classifier
Conditioning Chat-GPT for Information Retrieval: The Unipa-GPT Case Study
Full text linkThis paper illustrates the architecture and training of Unipa-GPT, a Large Language Model based chatbot developed for assisting students in choosing a bachelor/master degree course at the University of Palermo. Unipa-GPT relies on gpt-3.5-turbo, it was presented in the context of the European Researchers' Night SHARPER event. In our experiments we adopted both the Retrieval Augmented Generation (RAG) approach and fine-tuning to develop the system. The whole architecture of Unipa-GPT is presented, both the RAG and the fine-tuned systems are compared, and a brief discussion on their performance is reported
NL4AI 2023: Overview of the Seventh Workshop on Natural Language for Artificial Intelligence (NL4AI 2023)
No full textThe Natural Language for Artificial Intelligence (NL4AI) workshop serves as a platform to explore the area situated at the intersection between Natural Language Processing (NLP) and Artificial Intelligence (AI), with a special emphasis on recent activities carried out in both fields in Italy. The seventh edition of the workshop set a new record with 23 submissions, of which 18 were accepted. The submissions span a broad spectrum of topics, encompassing foundational NLP research, applied NLP, and works that bridge the realms of NLP and AI. Notably, this edition exhibited a growing international presence, featuring contributions from authors representing 9 countries. The submissions also reflect a diversity of languages (e.g., English, French, Italian) and modalities (e.g., text, vision), underscoring the workshop's commitment to inclusivity and comprehensive exploration
- …
