Multi-label classification assigns multiple labels to each document concurrently. Many real-world problems tend employ high-dimensional label spaces, which can be naturally structured in a hierarchy. In this type of problem, instance may belong and are organized hierarchical structure. It presents more complex problem than flat classification, given that the algorithm has take into account relationships between able predict for same instance. Few studies have investigated multi-label text...

Sentiment analysis became a very motivating area in both academic and industrial fields due to the exponential increase of online published reviews recommendations. To solve problem analysing classifying those recommendations, several techniques have been proposed. Lately, deep neural networks showed promising outcomes sentiment analysis. The growing number Arab users on Internet along with increasing amount Arabic comments encouraged researchers apply learning analyse them. This article is...

This paper focuses on clustering methods for content-based image retrieval CBIR. Hierarchical are a way to investigate grouping in data, simultaneously over variety of scales, by creating cluster tree. Traditionally, these group the objects into binary hierarchical Our main contribution is proposal new divisive hierarchy that based construction non-binary Each node can have more than two clusters detecting better m classes . To determine how divide nodes tree nodes, we use K-means...

The retrieving method proposed in this paper utilizes the fusion of images' multimodal information (textual and visual) which is a recent trend image retrieval researches. It combines two different data mining techniques to retrieve semantically related images: clustering association rules algorithm. semantic constructed at offline phase where are discovered between text clusters visual images use it later online phase. experiment was conducted on more than 54,500 ImageCLEF 2011 Wikipedia...

There is a massive growth of text documents on the web. This led to increasing need for methods that can organize and classify electronic (instances) automati-cally. Multi-label classification task widely used in real-world problems it has been applied di˙erent applications. It assigns multiple labels each document simultaneously. Few insuÿcient research studies have investigated multi-label problem Arabic language. Therefore, this survey paper aims present an extensive review existing...

Automatic image annotation (AIA) has been adopted in different applications such as retrieval and classification. Deep Learning is used AIA to extract features then convert these into text descriptions labels. However, conventional models that employ deep learning methods suffer from various shortcomings, poor performance. This work proposes an model based on convolutional neural networks (CNNs), generative adversarial (GANs), transfer learning. GANs have attracted a lot of interest because...

Nowadays, access to Arabic historical documents has become easier and faster due the availability of digital copies manuscripts. However, dealing with them manually remains difficult costly when indexing, searching, or analyzing data. This paper proposes a deep learning-based word spotting method predict labels. A comparative study was performed by considering several Convolutional Neural Network (CNN) architectures transfer learning pre-trained models. Also, two different labels are used;...

Effective and fast retrieval of images from image datasets is not an easy task, especially with the continuous growth digital added everyday by used to web. Automatic annotation approach that has been proposed facilitate semantically related a query image. A multimodal method in this paper. The goal benefit visual features extracted their associated user tags. relies on clustering regroup text into clusters association rules mining generate associate clusters. In experimental evaluation, two...

Quadriplegic people are unable to use mobile devices without the aid of other persons which can be devastating for them both socially and economically. This has motivated many researchers propose hardware software solutions that operate as intermediates between impaired users their devices: accessibility switches, joysticks head movements. However, efficiency these tools is limited in some conditions. To alleviate this problem, we exploit electroencephalographic signals captured via an...

Content-based image retrieval (CBIR) has been one of the most important research areas in computer vision. It is a widely used method for searching images huge databases. In this paper we present CBIR system called NOHIS-Search. The based on indexing technique NOHIS-tree. two phases are described and performance illustrated with database ImagEval. NOHIS-Search was compared to other systems; first that using PDDP algorithm second sequential search. Results show outperforms systems.

In Content-Based Image Retrieval systems it is important to use an efficient indexing technique in order perform and accelerate the search huge databases. The used should also support high dimensions of image features. this paper we present hierarchical index NOHIS-tree (Non Overlapping Hierarchical Index Structure) when scale up very large We a study influence clustering on time. performance test results show that performs better than SR-tree. Tests keeps its performances dimensional...

Content-based image retrieval (CBIR) has been one of the most important research areas in computer vision. It is a widely used method for searching images huge databases. In this paper we present CBIR system called NOHIS-Search. The based on indexing technique NOHIS-tree. two phases are described and performance illustrated with database ImagEval. NOHIS-Search was compared to other systems; first that using PDDP algorithm second sequential search. Results show outperforms systems.

A large scale of data is posted every day on the social media platforms. Sentiment analysis classification one useful techniques to extract information from those data. To train sentiment models there a need for labeled data, and it challenging issues. The available datasets are collected manually by humans which time-consuming process. identification machine learning classifier that provides best performance second issue in analysis. new hybrid model proposed this paper. It relies use...