Search

Currently, there is a lack of theoretical studies that approach classification and machine learning from a general perspective, without focusing on specific applications. While these studies tackle real problems like combining low-level and high-level descriptors, they fail to offer a comprehensive approach that effectively addresses the combination issue in all cases. Consequently, several questions arise when implementing such an approach, including: (i) how to model the combination of information from low-level and high-level descriptors; and (ii) how to evaluate the method's robustness across different applications. As these questions remain open and challenging, this study aims to provide answers. In this paper, we propose a new framework called AVR: "Advancing Video Retrieval ", which consists of three subsystems and is based on an optimized combination of low-level and high-level descriptors to enhance data retrieval accuracy. Performance analysis demonstrates that our AVR system significantly improves accuracy metrics compared to existing systems. For instance, using Corel's dataset, AVR increases average accuracy from 79.5% to 98.12% for the Beach category and from 77.75% to 98.45% for the Mountain category, surpassing the performance of our previous ISE system [38]. Furthermore, AVR achieves an average accuracy improvement of 99.32%, compared to 91.76% for ISE. Our system also outperforms our previous works, VINAS [40] and VISEN [120] systems, in terms of concept detection. For instance, the Car_Racing concept's value increases from 0.03 to 0.45 with AVR, leading to improved search results using the TRECVID dataset and a 98.41% average accuracy improvement compared to VINAS (85%) and VISEN (88%). [ABSTRACT FROM AUTHOR]

The objective of this research was to design a method to assign topics to claims debunked by fact-checking agencies. During the fact-checking process, access to more structured knowledge is necessary; therefore, we aim to describe topics with semantic vocabulary. Classification of topics should go beyond simple connotations like instance-class and rather reflect broader phenomena that are recognized by fact checkers. The assignment of semantic entities is also crucial for the automatic verification of facts using the underlying knowledge graphs. Our method is based on sentence embeddings, various clustering methods (HDBSCAN, UMAP, K-means), semantic entity matching, and terms importance assessment based on TF-IDF. We represent our topics in semantic space using Wikidata Q-ids, DBpedia, Wikipedia topics, YAGO, and other relevant ontologies. Such an approach based on semantic entities also supports hierarchical navigation within topics. For evaluation, we compare topic modeling results with claims already tagged by fact checkers. The work presented in this paper is useful for researchers and practitioners interested in semantic topic modeling of fake news narratives. [ABSTRACT FROM AUTHOR]

Recent technological advancements have led to a significant increase in the quantity and accessibility of videos. The decrease in video acquisition costs and the increase in memory capacity have made it possible to store large video collections in computer systems. To effectively exploit these collections, it is crucial to have tools that facilitate access and management. In this paper, we present a multimedia retrieval approach that prioritizes the user's needs by starting with a text-based query. The approach consists of two main parts: (i) a new multi-level and deep-semantic video classification indexing method, and (ii) a query expansion mechanism and relevance feedback system to improve the results based on the user's feedback. Our contribution is demonstrated through the implementation of the Deep-VISEN prototype and experiments on a collection of 2700 videos and 62838 images. The results show that our algorithm is effective and precise. [ABSTRACT FROM AUTHOR]

Huge quantities of audio and video material are available at universities and teaching institutions, but their use can be limited because of the lack of intelligent search tools. This paper describes a possible way to set up an indexing scheme that offers a smart search modality, that combines semantic analysis of video/audio transcripts with the exact time positioning of uttered words. The proposal leverages NLP methods for topic modeling with lexical analysis of lessons' transcripts and builds a semantic hierarchical index into the corpus of lessons analyzed. Moreover, using abstracting summarization, the system can offer short summaries on the subject semantically implied by the search carried out. [ABSTRACT FROM AUTHOR]

Biomedical experts are facing challenges in keeping up with the vast amount of biomedical knowledge published daily. With millions of citations added to databases like MEDLINE/PubMed each year, efficiently accessing relevant information becomes crucial. Traditional term-based searches may lead to irrelevant or missed documents due to homonyms, synonyms, abbreviations, or term mismatch. To address this, semantic search approaches employing predefined concepts with associated synonyms and relations have been used to expand query terms and improve information retrieval. The National Library of Medicine (NLM) plays a significant role in this area, indexing citations in the MEDLINE database with topic descriptors from the Medical Subject Headings (MeSH) thesaurus, enabling advanced semantic search strategies to retrieve relevant citations, despite synonymy, and polysemy of biomedical terms. Over time, advancements in semantic indexing have been made, with Machine Learning facilitating the transition from manual to automatic semantic indexing in the biomedical literature. The paper highlights the journey of this transition, starting with manual semantic indexing and the initial efforts toward automatic indexing. The BioASQ challenge has served as a catalyst in revolutionizing the domain of semantic indexing, further pushing the boundaries of efficient knowledge retrieval in the biomedical field.

With the extensive availability of technological systems all over the world and the increasing diffusion of information and documents by users, especially for the Arabic population, the development of semantic plagiarism detection systems has become essential due to its importance for protecting the rights of authors. Developing such a system that can covers the content of all Arabic documents using ontology as a semantic resource would be a complex and time-consuming task and would require intelligent natural language-processing capabilities.In the context of Arabic plagiarism detection systems using an Arabic ontology and permitting these systems to support semantic representation to more efficiently verify the originality of the research and meet the needs of researchers, this paper presents a novel approach for addressing plagiarism named Multi-agents Indexing System. The proposed system is composed of three phases: (1) natural language processing phase, (2) indexing phase and (3) evaluation phase.Our experimentations are based on the training dataset released for the AraPlagDet curpus. The obtained results indicated that the proposed system has improved the performance of plagiarism detection in Arabic documents with semantic indexing and mutli-agents system.

This study aims to develop a Natural Language Processing based decision support system built from a repository of knowledge drawn from referral letters written between primary care doctors and specialist medical consultants. The developed system translates pre-processed referral letters into a semantic matrix of document vectors and a set of vocabulary features, based solely on the words used within each referral letter. The system applies a one-versus-rest heuristic using a Support Vector Machine (SVM) to convert a multinomial classification problem into individual binary classifications. Each document is matched to its probabilistic best fit specialism. The National Health Service Wales sourced 111,700 examples. Accuracy of 91.8% against 29 medical specialities is achieved. Accuracy increases to 97.4% and 99%, respectively, when also including one or two nearest neighbours to the best fit, providing a basis for informing the decision making of a medical professional. The study demonstrates the efficacy of using referral letters to allow or classification into specialisms and subsequent allocation of specialist care. The approach taken in this study does not require added ontologies and is readily extendable. The system offers support to medical professionals, particularly within training scenarios or where access to opinion may be in short supply. [ABSTRACT FROM AUTHOR]

In this paper, we introduce a multi-label lazy learning approach to deal with automatic semantic indexing in large document collections in the presence of complex and structured label vocabularies with high inter-label correlation. The proposed method is an evolution of the traditional k-Nearest Neighbors algorithm which uses a large autoencoder trained to map the large label space to a reduced size latent space and to regenerate the predicted labels from this latent space. We have evaluated our proposal in a large portion of the MEDLINE biomedical document collection which uses the Medical Subject Headings (MeSH) thesaurus as a controlled vocabulary. In our experiments we propose and evaluate several document representation approaches and different label autoencoder configurations. [ABSTRACT FROM AUTHOR]

With the extensive availability of technological systems all over the world and the increasing diffusion of information and documents by users, especially for the Arabic population, the development of semantic plagiarism detection systems has become essential due to its importance for protecting the rights of authors. Developing such a system that can covers the content of all Arabic documents using ontology as a semantic resource would be a complex and time-consuming task and would require intelligent natural language-processing capabilities. In the context of Arabic plagiarism detection systems using an Arabic ontology and permitting these systems to support semantic representation to more efficiently verify the originality of the research and meet the needs of researchers, this paper presents a novel approach for addressing plagiarism named Multi-agents Indexing System. The proposed system is composed of three phases: (1) natural language processing phase, (2) indexing phase and (3) evaluation phase. Our experimentations are based on the training dataset released for the AraPlagDet curpus. The obtained results indicated that the proposed system has improved the performance of plagiarism detection in Arabic documents with semantic indexing and mutli-agents system. [ABSTRACT FROM AUTHOR]

One of the most challenging aspects of developing information systems is the processing and management of large volumes of information. One way to overcome this problem is to implement efficient data indexing and classification systems. As large volumes of generated data comprise of non-structured textual data, developing text processing, management and indexing frameworks can play an important role in providing users with accurate information according to their preferences. In this paper, a novel method of semantic information processing, management and indexing is introduced. The main goals of this study is to integrate structured knowledge of ontology and Knowledge Bases (KBs) in the core components of the method, to enrich the contents of the documents, to have multi-level semantic network representation of textual resources, to introduce a hybrid weighting schema (salient score) and finally to propose a hybrid method of semantic similarity computation. The structured knowledge of ontology and KBs are integrated from all aspects of the proposed method. The obtained results indicate the accuracy and optimal performance of the proposed framework. The obtained results suggest that using knowledge-based models leads to higher performance and accuracy in identifying and classifying documents according to user preferences; however, if learning-based models are not provided with sufficient amount of training data, they cannot yield satisfying results. The results also demonstrate that the complete integration of ontology and KBs in information systems can significantly contribute to a better representation of documents and evidently superior functionality of information processing, management and indexing systems.

Huge quantities of audio and video material are available at universities and teaching institutions, but their use can be limited because of the lack of intelligent search tools. This paper describes a possible way to set up an indexing scheme that offers a smart search modality, that combines semantic analysis of video/audio transcripts with the exact time positioning of uttered words. The proposal leverages NLP methods for topic modeling with lexical analysis of lessons’ transcripts and builds a semantic hierarchical index into the corpus of lessons analyzed. Moreover, using abstracting summarization, the system can offer short summaries on the subject semantically implied by the search carried out.

The Medical Subject Headings (MeSH) thesaurus is a controlled vocabulary widely used in biomedical knowledge systems, particularly for semantic indexing of scientific literature. As the MeSH hierarchy evolves through annual version updates, some new descriptors are introduced that were not previously available. This paper explores the conceptual provenance of these new descriptors. In particular, we investigate whether such new descriptors have been previously covered by older descriptors and what is their current relation to them. To this end, we propose a framework to categorize new descriptors based on their current relation to older descriptors. Based on the proposed classification scheme, we quantify, analyze, and present the different types of new descriptors introduced in MeSH during the last fifteen years. The results show that only about 25% of new MeSH descriptors correspond to new emerging concepts, whereas the rest were previously covered by one or more existing descriptors, either implicitly or explicitly. Most of them were covered by a single existing descriptor and they usually end up as descendants of it in the current hierarchy, gradually leading toward a more fine-grained MeSH vocabulary. These insights about the dynamics of the thesaurus are useful for the retrospective study of scientific articles annotated with MeSH, but could also be used to inform the policy of updating the thesaurus in the future. [ABSTRACT FROM AUTHOR]

In data science, there is a lot of interest in Data Analytics and Machine Learning. Large volumes of domain-specific data are being collected by government agencies and private organisations alike, and this data may provide useful information on national security, cyber security, fraud detection, and marketing strategies. Google and Microsoft, for example, examine vast amounts of data for business assessments and choices that have an effect on current and future technological developments, microsoft also does this. Through a hierarchical learning process, machine learning systems extract complicated high-level abstractions as data representations. At each step of the hierarchy, complicated abstractions are learnt through building on the simpler abstraction defined in the previous step. Machine Learning is used in Data Analytics because it is capable of analyzing and learning from vast volumes of unlabeled and uncategorized raw data. As a result, it's a useful tool for data analytics. Our study examines how Machine Learning may be used to tackle some of the most difficult Data Analytics problems, such as extracting complex patterns from large volumes of data, Ontology Indexes, Data Tagging, and Fast Information Retrieval. Streaming data, high-dimensional data scalability, and distributed computing are all topics of Machine Learning research that need to be explored further in order to address particular Data Analytics challenges. We also delve into these subjects in depth. Finally, we create data sampling criteria and domain adaption modelling, as well as requirements for the generation of appropriate data abstractions, to provide new insights into critical future endeavors. [ABSTRACT FROM AUTHOR]

Quran is the most significant religious document in the Arabic language in Islamic law. Several Quranic search engines have been designed and widely used for the past two decades. However, these search engines have certain limitations. For example, in many cases, the search is unable to retrieve relevant verses because it is based on keywords or root search and does not rely on the semantic relation between words in the query. The main objective of the present paper is to design a semantic search engine based on ontology as an index. In our work, we focus on creating a new ontology for Quranic document based on a set of useful words extracted from the Quranic Earab book with grammatical functions that serve as concepts. This ontology will be used as an index in information retrieval. The main idea is to create links between the existing Quranic words in the same verse, which will be used with user's query to find the desired verses. We developed a graphical user interface with free and multiple inputs that convert users' Arabic queries into SPARQL queries and then retrieve relevant verses from the ontology. The obtained results show that our proposal provides heightened precision and recall compared to other search engines. [ABSTRACT FROM AUTHOR]

In traditional word-based information retrieval systems, a document is considered a set of words representing graphs without semantics. In this paper, we focus on enriching the similarity measure by using synonymy and performance evaluation of semantic indexing approaches to a document corpus. We will also present comparisons showing that the use of synonymy with Leacock and Chodorow measures increases the semantic similarity that makes research more efficient. [ABSTRACT FROM AUTHOR]

Abstract Background In this paper we present the approach that we employed to deal with large scale multi-label semantic indexing of biomedical papers. This work was mainly implemented within the context of the BioASQ challenge (2013–2017), a challenge concerned with biomedical semantic indexing and question answering. Methods Our main contribution is a MUlti-Label Ensemble method (MULE) that incorporates a McNemar statistical significance test in order to validate the combination of the constituent machine learning algorithms. Some secondary contributions include a study on the temporal aspects of the BioASQ corpus (observations apply also to the BioASQ’s super-set, the PubMed articles collection) and the proper parametrization of the algorithms used to deal with this challenging classification task. Results The ensemble method that we developed is compared to other approaches in experimental scenarios with subsets of the BioASQ corpus giving positive results. In our participation in the BioASQ challenge we obtained the first place in 2013 and the second place in the four following years, steadily outperforming MTI, the indexing system of the National Library of Medicine (NLM). Conclusions The results of our experimental comparisons, suggest that employing a statistical significance test to validate the ensemble method’s choices, is the optimal approach for ensembling multi-label classifiers, especially in contexts with many rare labels.

The massive amount of digital content generated daily in the modern world has created the need for an image retrieval system built on image analysis via image processing and machine learning, therefore this study explains the role of machine learning in bridging the semantic gap in content‐based image retrieval, proposes an automatic image annotation framework, in which training images are obtained from social media, and semantic indexing is achieved using a combination of supervised and unsupervised machine learning. Furthermore, the study also highlights the need for continuous vocabulary improvement for optimum system performance and recommends hardware implementation of machine learning algorithms to ensure high overall speed of image retrieval systems. [ABSTRACT FROM AUTHOR]

The classification of natural language texts has gained a growing importance in many real world applications due to its significant implications in relation to crucial tasks, such as Information Retrieval, Question Answering, Text Summarization, Natural Language Understanding. In this paper we present an analysis of a Deep Learning architecture devoted to text classification, considering the extreme multi-class and multi-label text classification problem, when a hierarchical label set is defined. The paper presents a methodology named Hierarchical Label Set Expansion (HLSE), used to regularize the data labels, and an analysis of the impact of different Word Embedding (WE) models that explicitly incorporate grammatical and syntactic features. We evaluate the aforementioned methodologies on the PubMed scientific articles collection, where a multi-class and multi-label text classification problem is defined with the Medical Subject Headings (MeSH) label set, a hierarchical set of 27,775 classes. The experimental assessment proves the usefulness of the proposed HLSE methodology and also provides some interesting results relating to the impact of different uses and combinations of WE models as input to the neural network in this kind of application. • Deep Neural Network architecture for extreme multilabel text classification. • Multi-label classification problem with a huge label space hierarchically organized. • Comparison among different word-embeddings methods for text representation. • Definition of a method for label set expansion exploiting the label hierarchy. • Experimental assessment based on flat and hierarchical measures. [ABSTRACT FROM AUTHOR]

In traditional word-based information retrieval systems, a document is considered a set of words representing graphs without semantics. In this paper, we focus on enriching the similarity measure by using synonymy and performance evaluation of semantic indexing approaches to a document corpus. We will also present comparisons showing that the use of synonymy with Leacock and Chodorow measures increases the semantic similarity that makes research more efficient. [ABSTRACT FROM AUTHOR]

To cope with the growing number of multimedia assets on smartphones and social media, an integrated approach for semantic indexing and retrieval is required. Here, we introduce a generic framework to fuse existing image and video analysis tools and algorithms into a unified semantic annotation, indexing and retrieval model resulting in a multimedia feature vector graph representing various levels of media content, media structures and media features. Utilizing artificial intelligence (AI) and machine learning (ML), these feature representations can provide accurate semantic indexing and retrieval. Here, we provide an overview of the generic multimedia analysis framework (GMAF) and the definition of a multimedia feature vector graph framework (MMFVGF). We also introduce AI4MMRA to detect differences, enhance semantics and refine weights in the feature vector graph. To address particular requirements on smartphones, we introduce an algorithm for fast indexing and retrieval of graph structures. Experiments to prove efficiency, effectiveness and quality of the algorithm are included. All in all, we describe a solution for highly flexible semantic indexing and retrieval that offers unique potential for applications such as social media or local applications on smartphones.