No abstract

This paper explores how intent classification can be improved by representing the class labels not as a discrete set of symbols but as a space where the word graphs associated to each class are mapped using typical graph embedding techniques. The approach, inspired by a previous algorithm used for an inverse dictionary task, allows the classification algorithm to take in account inter-class similarities provided by the repeated occurrence of some words in the training examples of the different classes. The classification is carried out by mapping text embeddings to the word graph embeddings of the classes. Focusing solely on improving the representation of the class label set, we show in experiments conducted in both private and public intent classification datasets, that better detection of out-of-scope examples (OOS) is achieved and, as a consequence, that the overall accuracy of intent classification is also improved. In particular, using the recently-released Larson dataset, an error of about 9.9% has been achieved for OOS detection, beating the previous state-of-the-art result by more than 31 percentage points.

No abstract

No abstract

The coronavirus pandemic accentuated the need for molecular diagnostic tests. A technique highly used to this end is the Polymerase Chain Reaction (PCR)—a sensitive and specific technique commonly used as the gold standard for molecular diagnostics. However, it demands highly trained personnel and high-maintenance equipment and is relatively time-consuming. An alternative is the Loop-Mediated Isothermal Amplification (LAMP) technique, which doesn’t need sample purification or expensive equipment, and is similar to PCR when compared in sensitivity and specificity. In this paper, we developed an optimized colorimetric Reverse Transcriptase Loop-Mediated Isothermal Amplification (RT-LAMP) Point-of-Care test using a portable device to diagnose COVID-19. Variables such as concentration of primers, magnesium sulfate, betaine, hydrochloride guanidine, Bst, and temperature of the reactions were tested. We also created a pipetting quality control system—using a combination of dyes—to avoid false negatives due to a lack of samples added to the reaction test tube. Mineral oil was incorporated in the composition of the RT-LAMP reactions to avoid evaporation when a heating lid isn't available. The final RT-LAMP test is tenfold more sensitive when compared to the WarmStart Colorimetric Master mix from New England Biolabs with a sensitivity of 5 copies per μL.