A New Approach for Video Concept Detection Based on User Comments

Video concept detection means describing a video with semantic concepts that correspond to the content of the video. The concepts help to retrieve video quickly. These semantic concepts describe high-level elements that depict the key information present in the content. In recent years, many efforts have been done to automate this task because the manual solution is time-consuming. Nowadays, videos come with comments. Therefore, in addition to the content of the videos, the comments should be analyzed because they contain valuable data that help to retrieve videos. This paper focused especially on videos shared on social media. The specificity of these videos was the presence of massive comments. This paper attempted to exploit comments by extracting concepts from them. This would support the research effort that works only on the visual content. Natural language processing techniques were used to analyze comments and to filter words to retain only the ones that could be considered as concepts. The proposed approach was tested on YouTube videos. The results demonstrated that the proposed approach was able to extract accurate data and concepts from the comments that could be used to ease the retrieval of videos. The findings supported the research effort of working on the visual and audio contents of the videos.

LLOD-Driven Bilingual Word Embeddings Rivaling Cross-Lingual Transformers in Quality of Life Concept Detection from French Online Health Communities

We describe the use of Linguistic Linked Open Data (LLOD) to support a cross-lingual transfer framework for concept detection in online health communities. Our goal is to develop multilingual text analytics as an enabler for analyzing health-related quality of life (HRQoL) from self-reported patient narratives. The framework capitalizes on supervised cross-lingual projection methods, so that labeled training data for a source language are sufficient and are not needed for target languages. Cross-lingual supervision is provided by LLOD lexical resources to learn bilingual word embeddings that are simultaneously tuned to represent an inventory of HRQoL concepts based on the World Health Organization’s quality of life surveys (WHOQOL). We demonstrate that lexicon induction from LLOD resources is a powerful method that yields rich and informative lexical resources for the cross-lingual concept detection task which can outperform existing domain-specific lexica. Furthermore, in a comparative evaluation we find that our models based on bilingual word embeddings exhibit a high degree of complementarity with an approach that integrates machine translation and rule-based extraction algorithms. In a combined configuration, our models rival the performance of state-of-the-art cross-lingual transformers, despite being of considerably lower model complexity.

Multimodal Video Indexing (MVI): A New Method Based on Machine Learning and Semi-Automatic Annotation on Large Video Collections

Indexing video by the concept is one of the most appropriate solutions for such problems. It is based on an association between a concept and its corresponding visual sound, or textual features. This kind of association is not a trivial task. It requires knowledge about the concept and its context. In this paper, we investigate a new concept detection approach to improve the performance of content-based multimedia documents retrieval systems. To achieve this goal, we are going to tackle the problem from different plans and make four contributions at various stages of the indexing process. We propose a new method for multimodal indexation based on (i) a new weakly supervised semi-automatic method based on the genetic algorithm (ii) the detection of concepts from the text in the videos (iii) the enrichment of the basic concepts thanks to the usage of our method DCM. Subsequently, the semantic and enriched concepts allow a better multimodal indexation and the construction of an ontology. Finally, the different contributions are tested and evaluated on a large dataset (TRECVID 2015).

Detection of Alzheimer's associated microRNAs using a DNA-based smart reagent

Alzheimer's disease (AD) is the most common neurodegenerative disorder, with significant research efforts devoted to identifying new biomarkers for clinical diagnosis and treatment. MicroRNAs have emerged as likely disease regulators and biomarkers for AD, now implicated as having roles in several biological processes related to progression of the disease. In this work, we use the miRacles assay (microRNA activated conditional looping of engineered switches) for single-step detection of AD-related microRNAs. The technology is based on conformationally responsive DNA nanoswitches that loop upon recognition of a target microRNA and report their on/off status through an electrophoretic readout. Unlike many other methods, our approach directly detects native microRNAs without amplification or labeling, eliminating the need for expensive enzymes, reagents, and equipment. We used this assay to screen for AD-related microRNAs, demonstrate specificity within a microRNA family, sensitivity of ~ 8 fM, and multiplexing capability to simultaneously detect four microRNA targets. Toward clinical use, we provide proof-of-concept detection and quantifiable dysregulation of specific microRNAs from total RNA extracts derived from healthy and AD brain samples. In the context of AD, this "smart reagent" could facilitate biomarker discovery, accelerate efforts to understand the role of microRNAs in AD, and have clinical potential as a diagnostic or monitoring tool for validated biomarkers.

A Paper-Based Colorimetric Aptasensor for the Detection of Gentamicin

Antibiotics are classes of antimicrobial substances that are administered widely in the field of veterinary science to promote animal health and feed efficiency. Cattle-administered antibiotics hold a risk of passing active residues to milk, during the milking process. This becomes a public health concern as these residues can cause severe allergic reactions to sensitive groups and considerable economic losses to the farmer. Hence, to ensure that the produced milk is safe to consume and adheres to permissible limits, an on-farm quick and reliable test is essential. This study illustrates the design and development of a microfluidic paper biosensor as a proof-of-concept detection system for gentamicin in milk. Localized surface plasmon resonance (LSPR) properties of gold nanoparticles have been explored to provide the user a visual feedback on the test, which was also corroborated by RGB analysis performed using Image J. The assay involves the use of a short stretch of single stranded DNA, called aptamer, which is very specific to the gentamicin present in the milk sample. The camera-based LOD for the fabricated paper device for milk samples spiked with gentamicin was calculated to be 300 nM, with a reaction time of 2 min.