Sylvie Ranwez

Sylvie Ranwez
IMT Mines Alès | EMA · EuroMov Digital Health in Motion

PhD

About

92
Publications
18,417
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
1,180
Citations
Citations since 2016
21 Research Items
582 Citations
2016201720182019202020212022020406080100
2016201720182019202020212022020406080100
2016201720182019202020212022020406080100
2016201720182019202020212022020406080100
Introduction
Recommander system using semantics, Mouvement annotation (multi modality)
Additional affiliations
July 2017 - present
IMT Mines Alès
Position
  • Professor
Education
September 1997 - December 2000
Montpellier University
Field of study
  • Computer science

Publications

Publications (92)
Article
Full-text available
Leveraging knowledge graphs for post-hoc recommendation explanations has been investigated in recent years. Existing approaches rely mainly on the overlap properties (encoded by knowledge graphs) that characterize both user liked items and the recommended ones. These approaches, however, do not fully leverage the property hierarchy of knowledge gra...
Article
Full-text available
The diversity of the item list suggested by recommender systems has been proven to impact user satisfaction significantly. Most of the existing diversity optimization approaches re-rank the list of candidate items during a post-processing step. However, the diversity level of the candidate list strongly depends on the recommender system used. Hence...
Article
Full-text available
Recommender systems aim to provide users with a selection of items, based on predicting their preferences for items they have not yet rated, thus helping them filter out irrelevant ones from a large product catalogue. Collaborative filtering is a widely used mechanism to predict a particular user’s interest in a given item, based on feedback from n...
Research
Il vous est arrivé de rester perplexe devant une recommandation de Netflix, Youtube ou Amazon Prime... Nous aussi ! En moins de 10 minutes vous pouvez aider à améliorer les choses : sélectionnez quelques films que vous avez particulièrement aimés et évaluez les recommandations qui vous sont faites (et les explications qui vont avec). Vous contribu...
Preprint
Full-text available
Patent analysis and mining are time-consuming and costly processes for companies, but nevertheless essential if they are willing to remain competitive. To face the overload induced by numerous patents, the idea is to automatically filter them, bringing only few to read to experts. This paper reports a successful application of fine-tuning and retra...
Preprint
Full-text available
This study presents a large scale benchmarking on cloud based Speech-To-Text systems: {Google Cloud Speech-To-Text}, {Microsoft Azure Cognitive Services}, {Amazon Transcribe}, {IBM Watson Speech to Text}. For each systems, 40158 clean and noisy speech files about 101 hours are tested. Effect of background noise on STT quality is also evaluated with...
Article
Data veracity is one of the main issues regarding Web data. Truth Discovery models can be used to assess it by estimating value confidence and source trustworthiness through analysis of claims on the same real-world entities provided by different sources. Many studies have been conducted in this domain. True values selected by most models have the...
Article
Data veracity is one of the main issues regarding web data. Facing fake news proliferation and disinformation dangers, Truth Discovery models can be used to assess this veracity by estimating value confidence and source trustworthiness through analysis of claims on the same real-world entities provided by different sources. This treatment is crucia...
Article
Full-text available
This article introduces an automated knowledge inference approach taking advantage of relationships extracted from texts. It is based on a novel framework making possible to exploit (i) a generated partial ordering of studied objects (e.g. noun phrases), and (ii) prior knowledge defined into ontologies. This framework is particularly suited for def...
Conference Paper
The main aim of truth-finding methods is to identify the most reliable and trustworthy data among a set of facts. Since existing methods assume a single true value, they cannot deal with numerous real-world use cases in which a set of true values exists for a given fact, even for functional predicate (e.g. Picasso is born in Màlaga and in Spain). T...
Conference Paper
Full-text available
Designing approaches able to automatically detect uncertain expressions within natural language is central to design efficient models based on text analysis, in particular in domains such as question-answering, approximate reasoning , knowledge-based population. This article proposes an overview of several contributions and classifications defining...
Article
Full-text available
La détection de l'incertitude dans le langage naturel est centrale pour le développe-ment de nombreux modèles exploitant l'analyse de textes e.g. questions-réponses, raisonnement approché, enrichissement de bases de connaissances. Après une synthèse des différentes classifications de l'incertitude et des méthodes de détection correspondantes, cet a...
Conference Paper
Full-text available
The need of indexing biomedical papers with the MeSH is incessantly growing and automated approaches are constantly evolving. Since 2013, the BioASQ challenge has been promoting those evolutions by proposing datasets and evaluation metrics. In this paper, we present our system, USI, and how we adapted it to participate to this challenge this year....
Article
Full-text available
Malgré leur volume important et leur accessibilité, de nombreuses données numériques ne peuvent être correctement exploitées car elles sont contenues dans des textes sous des formes peu ou pas structurées. L’extraction de relations est un processus qui rassemble des techniques pour extraire des entités et des relations à partir de textes, nous donn...
Book
Full-text available
Artificial Intelligence federates numerous scientific fields in the aim of developing machines able to assist human operators performing complex treatments -- most of which demand high cognitive skills (e.g. learning or decision processes). Central to this quest is to give machines the ability to estimate the likeness or similarity between things i...
Article
Full-text available
Background Semantic approaches such as concept-based information retrieval rely on a corpus in which resources are indexed by concepts belonging to a domain ontology. In order to keep such applications up-to-date, new entities need to be frequently annotated to enrich the corpus. However, this task is time-consuming and requires a high-level of exp...
Chapter
The capacity of assessing the similarity of objects or stimuli has long been characterized as a central component for establishing numerous cognitive processes. It is therefore not surprising that measures of similarity or distance play an important role in a large variety of treatments and algorithms, and are of particular interest for the develop...
Chapter
Back in the 60s, the quest for artificial intelligence (AI) had originally been motivated by the assumption that “[…] every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it […]” [McCarthy et al., 2006]. Even if this assumption has today proved to be pretenti...
Chapter
This chapter is dedicated to semantic measure evaluation and discusses in particular two important topics: (i) how to objectively evaluate measures and (ii) how to guide their selection with regard to specific needs. To tackle these central questions, we propose technical discussions on both methodological and practical aspects related to semantic...
Chapter
As we have seen, two main families of semantic measures can be distinguished: corpus-based measures, which take advantage of unstructured or semi-structured texts, and knowledge-based measures which rely on ontologies.
Conference Paper
Full-text available
Concept-based information retrieval is known to be a powerful and reliable process. It relies on a semantically annotated corpus, i.e. resources indexed by concepts organized within a domain ontology. The conception and enlargement of such index is a tedious task, which is often a bottleneck due to the lack of (semi-)automated solutions. In this pa...
Conference Paper
Full-text available
Knowledge-based semantic measures are cornerstone to exploit ontologies not only for exact inferences or retrieval processes, but also for data analyses and inexact searches. Abstract theoretical frameworks have recently been proposed in order to study the large diversity of measures available; they demonstrate that groups of measures are particula...
Conference Paper
Full-text available
Semantic similarity and relatedness are cornerstones of numerous treatments in which lexical units (e.g., terms, documents), concepts or instances have to be compared from texts or knowledge representation analysis. These semantic measures are central for NLP, information retrieval, sentiment analysis and approximate reasoning, to mention a few. In...
Article
Full-text available
The Semantic Measures Library and Toolkit are robust open source and easy to use software solutions dedicated to semantic measures. They can be used for large scale computations and analyses of semantic similarities between terms/concepts defined in terminologies and ontologies. The comparison of entities (e.g. genes) annotated by concepts is also...
Article
Full-text available
Semantic measures are today widely used to estimate the strength of the semantic relationship between elements of various types: units of language (e.g., words, sentences), concepts or even entities (e.g., documents, genes, geographical locations). They play an important role for the comparison these elements according to semantic proxies, texts an...
Conference Paper
Full-text available
Many applications take advantage of both ontologies and the Linked Data paradigm to characterize various kinds of resources. To fully exploit this knowledge, measures are used to estimate the relatedness of resources regarding their semantic characterization. Such semantic measures mainly focus on specific aspects of the semantic characterization (...
Conference Paper
Full-text available
Semantic Measures (SMs) are of critical importance in multiple treatments relying on ontologies. However, the improvement and use of SMs are currently hampered by the lack of a dedicated theoretical framework and an extensive generic software solution. To meet these needs, this paper introduces a unified theoretical framework of graph-based SMs, fr...
Article
Impliqué dans un processus décisionnel, l'opérateur humain est souvent confronté à un trop grand nombre d'informations, qu'il doit analyser, synthétiser et exploiter parfois dans un contexte de crise où le facteur temps est décisif. Il est alors nécessaire d'automatiser certaines tâches à haute valeur cognitive ajoutée pour optimiser ce processus d...
Chapter
Full-text available
The exponential growth of available electronic data is almost useless without efficient tools to retrieve the right information at the right time. It is now widely acknowledged that information retrieval systems need to take semantics into account to enhance the use of available information. However, there is still a gap between the amounts of rele...
Article
Ontologies are widely adopted in the biomedical domain to characterize various resources (e.g. diseases, drugs, scientific publications) with non-ambiguous meanings. By exploiting the structured knowledge that ontologies provide, a plethora of ad hoc and domain-specific semantic similarity measures have been defined over the last years. Nevertheles...
Article
Full-text available
Ontologies are successfully used as semantic guides when navigating through the huge and ever increasing quantity of digital documents. Nevertheless, the size of numerous domain ontologies tends to grow beyond the human capacity to grasp information. This growth is problematic for a lot of key applications that require user interactions such as doc...
Article
Full-text available
The least common ancestor on two vertices, denoted lca(x,y), is a well defined operation in a directed acyclic graph (dag) G. We introduce U lca (S), a natural extension of lca(x,y) for any set S of vertices. Given such a set S 0 , one can iterate S k+1 =U lca (S k ) in order to obtain an increasing set sequence. G being finite, this sequence has a...
Article
Full-text available
Les performances d'un système de recherche d'information (SRI) peuvent être dégradées en termes de précision du fait de la difficulté pour des utilisateurs à formuler précisément leurs besoins en information. La reformulation ou l'expansion de requêtes constitue une des réponses à ce problème dans le cadre des SRI. Dans cet article, nous proposons...
Article
Full-text available
During last decade, ontologies have been successfully used as semantic guidelines while navigating through huge and ever increasing quantity of digital documents. Nevertheless, the size of most ontologies, especially those shared and accepted as standards in a given domain, tends to grow beyond the human capacity to grasp information. This growth i...
Article
Full-text available
Background: Because of the increasing number of electronic resources, designing efficient tools to retrieve and exploit them is a major challenge. Some improvements have been offered by semantic Web technologies and applications based on domain ontologies. In life science, for instance, the Gene Ontology is widely exploited in genomic applications...
Conference Paper
Full-text available
Because of the increasing number of electronic data, designing efficient tools to retrieve and exploit documents is a major challenge. Current search engines suffer from two main drawbacks: there is limited interaction with the list of retrieved documents and no explanation for their adequacy to the query. Users may thus be confused by the selectio...
Article
Full-text available
Pour exploiter efficacement des corpus documentaires toujours plus volumineux, les moteurs de recherche doivent évoluer. Leurs limites actuelles concernent principalement le fait que la mesure de la pertinence d'un document par rapport à une requête est souvent non-explicite et que l'interaction avec la liste des réponses est limitée. Nous proposon...
Article
Full-text available
Social photos, which are taken during family events or parties, represent individuals or groups of people. We show in this paper how a Hasse diagram is an efficient visualization strategy for eliciting different groups and navigating through them. However, we do not limit this strategy to these traditional uses. Instead we show how it can also be u...
Article
Full-text available
Nous nommons ‘photos sociales' les photos qui sont prises lors d'événements familiaux ou de soirées entre amis et qui représentent des individus ou des groupes d'individus. Leur indexation consiste à repérer l'événement et les personnes présentes sur les photos. Dans cet article nous présentons une méthode et des outils pour faciliter cette tâche....
Article
The increasing size of indexed document sets that are digitally available emphasises the crucial need for more suitable representation tools than traditional textual lists of results. Many efforts have been made to develop graphical tools capable of providing both overall and local views of a collection when focusing on a particular subset of docum...
Conference Paper
Full-text available
The increasing size of structured data that are digitally available emphasizes the crucial need for more suitable representation tools than the traditional textual list of results. A suitable visual representation should both reflect the database's structure for navigation purpose and allow performing visual analytical tasks for knowledge extractio...
Article
Narrative abstraction consists in selecting and assembling meaningful events from an original set of related events. This acquisition of information hinges on several requirements. This paper deals with some of them, namely, the viewer's intention, the viewer's resource constraints, particularly the time constraint, and the narrative coherence. We...
Article
Full-text available
Si les connaissances disponibles en ligne sont massives, leur partage au sein d'une communauté requiert une structuration et des techniques adaptées. En particulier, leur exploitation serait facilitée par une visualisation intuitive et organisée. Nous proposons une méthode, un environnement logiciel, et des heuristiques pour créer des cartes auto-o...