Semi-supervised learning and bidirectional decoding for effective grammar correction in low-resource scenarios.

The correction of grammatical errors in natural language processing is a crucial task as it aims to enhance the accuracy and intelligibility of written language. However, developing a grammatical error correction (GEC) framework for low-resource languages presents significant challenges due to the lack of available training data. This article proposes a novel GEC framework for low-resource languages, using Arabic as a case study. To generate more training data, we propose a semi-supervised confusion method called the equal distribution of synthetic errors (EDSE), which generates a wide range of parallel training data. Additionally, this article addresses two limitations of the classical seq2seq GEC model, which are unbalanced outputs due to the unidirectional decoder and exposure bias during inference. To overcome these limitations, we apply a knowledge distillation technique from neural machine translation. This method utilizes two decoders, a forward decoder right-to-left and a backward decoder left-to-right, and measures their agreement using Kullback-Leibler divergence as a regularization term. The experimental results on two benchmarks demonstrate that our proposed framework outperforms the Transformer baseline and two widely used bidirectional decoding techniques, namely asynchronous and synchronous bidirectional decoding. Furthermore, the proposed framework reported the highest F1 score, and generating synthetic data using the equal distribution technique for syntactic errors resulted in a significant improvement in performance. These findings demonstrate the effectiveness of the proposed framework for improving grammatical error correction for low-resource languages, particularly for the Arabic language.

RFID sensor-tags feeding a context-aware rule-based healthcare monitoring system.

Along with the growing of the aging population and the necessity of efficient wellness systems, there is a mounting demand for new technological solutions able to support remote and proactive healthcare. An answer to this need could be provided by the joint use of the emerging Radio Frequency Identification (RFID) technologies and advanced software choices. This paper presents a proposal for a context-aware infrastructure for ubiquitous and pervasive monitoring of heterogeneous healthcare-related scenarios, fed by RFID-based wireless sensors nodes. The software framework is based on a general purpose architecture exploiting three key implementation choices: ontology representation, multi-agent paradigm and rule-based logic. From the hardware point of view, the sensing and gathering of context-data is demanded to a new Enhanced RFID Sensor-Tag. This new device, de facto, makes possible the easy integration between RFID and generic sensors, guaranteeing flexibility and preserving the benefits in terms of simplicity of use and low cost of UHF RFID technology. The system is very efficient and versatile and its customization to new scenarios requires a very reduced effort, substantially limited to the update/extension of the ontology codification. Its effectiveness is demonstrated by reporting both customization effort and performance results obtained from validation in two different healthcare monitoring contexts.