Abstract
Sentiment analysis is one of the most widely studied applications in NLP, but most work focuses on languages with large amounts of data. We introduce the first large-scale human-annotated Twitter sentiment dataset for the four most widely spoken languages in Nigeria-Hausa, Igbo, Nigerian-Pidgin, and Yoruba-consisting of around 30,000 annotated tweets per language, including a significant fraction of code-mixed tweets. We propose text collection, filtering, processing, and labeling methods that enable us to create datasets for these low-resource languages. We evaluate a range of pre-trained models and transfer strategies on the dataset. We find that language-specific models and language-adaptive fine-tuning generally perform best. We release the datasets, trained models, sentiment lexicons, and code to incentivize research on sentiment analysis in under-represented languages.

Abstract

Abstract
This paper reports the development of WalkSafe, a Smart Pedestrian Crossing solution, by a multinational and multidisciplinary team of students during the spring semester of 2020. The team was enrolled in the European Project Semester (EPS), a project-based capstone programme offered by Instituto Superior de Engenharia do Porto (ISEP). Motivated by the idea to reduce the number of pedestrians hit by cars at road crossings, and associated injuries and deaths, the team surveyed pedestrian behaviour to conclude that people often ignore pedestrian crossings. Thus, this project intended to motivate people to use pedestrian crossings, increasing the safety of both pedestrians and drivers. The proposed solution can be implemented on any pedestrian crossing and involves up to three components: (i) a box to be fixed on each side of pedestrian crossings with a radio-frequency identification reader as well as Bluetooth and Wi-Fi interfaces; (ii) a smartphone mobile app; and (iii) a personal bracelet for children and elderly, with a passive radio-frequency identification tag.

Abstract
Daily tasks like reading the news or just saying hello to a friend or family member is through technologies, online applications in the mobile devices. Being online is a natural way of life, but those how did not grow with these new technologies and the internet feel outcast from the new online world. Senior citizens are looking to get themselves an update, connect with this online world, to be closer to family and friend members. It is simple as that, but is it? No. Without the fundamental knowledge on how to be safe on online browsing or using social media applications are the personal data of senior citizen safe and their own life safe? No. Through serious games, one can help them to learn all the matters in digital skills and more importantly in online web browsing.

Abstract
Biogenic amines (BAs) are compounds found in a vast range of food products. In recent years, there has been a crescent awareness toward food safety, followed by an increase in food regulations. Long-period fiber gratings (LPFGs) coated with titanium dioxide (TiO2) were used to monitor the optical properties of a layer of poly(ethylene-co-vinyl acetate) (PEVA) doped with maleic anhydride (MA), which was polymerized on top of TiO2. This hydrophobic polymeric structure is permeable to BA, which causes a steady increase in its effective refractive index (RI) causing a wavelength shift in the coated LPFG attenuation band. LPFG wavelength shift was observed and measured for the monoamine tyramine (TYR), to the diamines, putrescine (PUT), cadaverine (CAD), histamine (HIS), and tryptamine (TRYP), and to the polyamines, spermidine (SPED), and spermine (SPEM). It was determined that, while PEVA-coated devices present a residual sensitivity to BA, the MA greatly increases it. In fact, for PEVA only coated LPFGs, the sensitivities of 1.45 +/- 0.11, 0.97 +/- 0.05, 0.46 +/- 0.08, and 0.94 +/- 0.09 nmM-1 for PUT, CAD, HIS, and TYR, respectively, were measured. However, for PEVA-doped MA-coated LPFGs, the sensitivities are 3.34 +/- 0.13, 3.06 +/- 0.11, 2.62 +/- 0.14, and 3.65 +/- 0.23 nmM-1 for PUT, CAD, HIS, and TYR, respectively. Thus, the RI of PEVA increases with BAs in- diffusion, and MA doping further enhances the PEVA sensitivity to BA. The proposed sensor is expected to play a part in the further development of a biosensor for the quantification of BA in real foodstuff, providing a methodology for quality control.

Abstract
In this paper, we study the implementation of a secure key distribution system based on an ultra-long fiber laser with a bi-directional erbium-doped fiber amplifier. The resilience of the system was tested against passive attacks from an eavesdropper. A similarity was observed in the spectra for both secure configurations of the system and no signature that would allow an eavesdropper to obtain the secure state of the system was observed during the state transitions.