INESC-ID RCAAP Portal
Not a member yet
15883 research outputs found
Sort by
Children's Voice Privacy: First Steps and Emerging Challenges
No full textThis study addresses the underrepresented area of children's voice privacy by establishing a baseline for voice anonymization techniques designed for adult speech applied to children's voices. The authors evaluate six anonymization methods across three children's datasets and provide both objective and subjective evaluations of their performance, revealing unique challenges inherent to children's voice anonymization. The findings highlight the need for tailored systems to address the characteristics of children's speech, underscoring the importance of prioritizing children's privacy in digital environments
Fusing Session-Typed Concurrent Programming into Functional Programming
No full textThis research paper introduces FuSes, a functional programming language that integrates session-typed concurrent process calculus, enabling safe and modular interoperability between pure functional programming and effectful concurrency. The authors provide a two-layered system with a functional layer generating process code and a message-passing process layer sending and receiving functional values. By extending a core functional language with primitives to run and observe process behavior, FuSes bridges the gap between functional and concurrent programming paradigms, facilitating type-safe metaprogramming and analysis of open session-typed process code
Portuguese Far-Right Discourse on Social Media: Insights from Topic Modeling
No full textThis study analyzes the social media discourse of CHEGA, Portugal's far-right party, using topic modeling to identify latent topics and discursive dynamics. The research reveals two main themes: ideological and public, and party-related, with a focus on identity, immigration, and security narratives that are strategically used for political mobilization during electoral cycles. By leveraging BERTopic, a robust topic-modeling technique, the study demonstrates how computational methods can uncover strategic communication patterns in far-right political discourse, shedding light on the digital ecosystem's impact on contemporary politics
