Oumaima Alaoui Ismaili's research while affiliated with Orange Labs and other places
What is this page?
This page lists the scientific contributions of an author, who either does not have a ResearchGate profile, or has not yet added these contributions to their profile.
It was automatically created by ResearchGate to create a record of this author's body of work. We create such pages to advance our goal of creating and maintaining the most comprehensive scientific repository possible. In doing so, we process publicly available (personal) data relating to the author as a member of the scientific community.
If you're a ResearchGate member, you can follow this page to keep up with this author's work.
If you are this author, and you don't want us to display this page anymore, please let us know.
It was automatically created by ResearchGate to create a record of this author's body of work. We create such pages to advance our goal of creating and maintaining the most comprehensive scientific repository possible. In doing so, we process publicly available (personal) data relating to the author as a member of the scientific community.
If you're a ResearchGate member, you can follow this page to keep up with this author's work.
If you are this author, and you don't want us to display this page anymore, please let us know.
Publications (22)
At the stage of identifying and modeling business processes (BP), traditional methods for collecting business expertise are time consuming since they mainly rely on physical communication. Moreover, there are some BP with no formal documentation and their execution relies on implicit knowledge of workers. Therefore, analyzing the logs data generate...
Supervised classification can be effective for prediction but sometimes weak on interpretability or explainability (XAI). Clustering, on the other hand, tends to isolate categories or profiles that can be meaningful but there is no guarantee that they are useful for labels prediction. Predictive clustering seeks to obtain the best of the two worlds...
Process mining aims at discovering different perspectives of business processes (BP) from event logs generated by BP management systems. However, BP can be entirely or partially performed outside such systems. Emails are widely used as an alternative tool to collaboratively perform BP tasks. Recently, there have been several initiatives to extend t...
Supervised classification can be effective for prediction but sometimes weak on interpretability or explainability (XAI). Clustering, on the other hand, tends to isolate categories or profiles that can be meaningful but there is no guarantee that they are useful for labels prediction. Predictive clustering seeks to obtain the best of the two worlds...
Significant research work has been conducted in the area of process mining leading to mature solutions for discovering knowledge from structured process event logs analysis. Recently, there have been several initiatives to extend the scope of these analysis to consider heterogeneous and unstructured data sources. More precisely, email analysis has...
Supervised classification can be effective for prediction but sometimes weak on interpretability or explainability (XAI). Clustering, on the other hand, tends to isolate categories or profiles that can be meaningful but there is no guarantee that they are useful for labels prediction. Predictive clustering seeks to obtain the best of the two worlds...
Dans le cadre du clustering prédictif, pour attribuer la classe aux groupes formés à la fin de la phase d’apprentissage, le vote majoritaire est la méthode communément utilisée. Cependant, cette approche comporte certaines limitations qui influent directement sur la qualité des résultats obtenus en termes de prédition. Pour surmonter ce problème, n...
L’algorithme des K-moyennes prédictives est un des algorithmes de clustering prédictif visant à décrire et à prédire d’une manière simultanée. Contrairement à la classification supervisée et au clustering traditionnel, la performance de ce type d’algorithme est étroitement liée à sa capacité à réaliser un bon compromis entre la description et la pr...
Le clustering prédictif est un nouvel aspect d’apprentissage supervisé dérivé du clustering standard. Les algorithmes appartenant à ce type de l’apprentissage cherchent à décrire et à prédire d’une manière simultanée. Il s’agit de découvrir la structure interne d’une variable cible. Puis munis de cette structure, de prédire la classe des nouvelles...
Over the last years, researchers have focused their attention on a new approach, supervised clustering, that combines the main characteristics of both traditional clustering
and supervised classification
tasks. Motivated by the importance of pre-processing approaches in the traditional clustering
context, this paper explores to what extent supervis...
Over the last years, researchers have focused their attention on a new approach, supervised clustering, that combines the main characteristics of both traditional clustering and supervised classification tasks. Motivated by the importance of initialization step in the traditional clustering context, this paper explores to what extent supervised ini...
Dans certains domaines applicatifs, la compréhension (la description) des résultats issus d’un classifieur est une condition aussi importante que sa performance prédictive. De ce fait, la qualité du classifieur réside donc dans sa capacité à fournir des résultats ayant de bonnes performances en prédiction tout en produisant simultanément des résult...
Over the last years, researchers have focused their
attention on a new approach, supervised clustering, that combines
the main characteristics of both traditional clustering and
supervised classification tasks. Motivated by the importance of
the initialization in the traditional clustering context, this paper
explores to what extent supervised init...
Over the last years, researchers focus their attention on a new approach that combines the main characteristics of both traditional clustering and supervised classification tasks. This new approach is called by supervised clustering. Motivated by the importance of pre-processing approaches in a traditional clustering context, we suppose that a supe...
This article proposes a supervised approach to evaluate the contribution of explanatory variables to a clustering. The main idea is to learn to predict the instance membership to the clusters using each individual variable. All variables are then sorted with respect to their predictive power, which is measured using two evaluation criteria, i.e. ac...
This article proposes a supervised approach to evaluate the contribution of explanatory variables to a clustering. The main idea is to learn to predict the instance membership to the clusters using each individual variable. All variables are then sorted with respect to their predictive power, which is measured using two evaluation criteria, i.e. ac...
Citations
... ) was published in the international conference on Enabling Technologies: Infrastructure for Collaborative Enterprises (WETICE)[30] and the Concurrency and Computation: Practice and Experience journal[32]. ...
![[object Object]](https://i1.rgstatic.net/ii/profile.image/933833566334984-1599654680929_Q64/Marwa-Elleuch.jpg)
![[object Object]](https://i1.rgstatic.net/ii/profile.image/272492417777716-1441978664004_Q64/Walid-Gaaloul.jpg)
... ) was published in the international conference on Enabling Technologies: Infrastructure for Collaborative Enterprises (WETICE)[30] and the Concurrency and Computation: Practice and Experience journal[32]. ...
... We show the results of two experiments carried out to prove the validity of the main design choices adopted in this metamodel. Some parts in this chapter were published in the International Conference on Services Computing (SCC) [28]. ...
![[object Object]](https://i1.rgstatic.net/ii/profile.image/318493520531462-1452946181049_Q64/Nour-Assy-2.jpg)
... To enable the discovery of multiple activities per email, proposed approaches relies on: (1) splitting emails into sentences and then, assigning these sentences to activities (using supervised or unsupervised learning techniques) [18,43,76,77,86,48], or (2) associating to each activity a set of patterns of words enabling its detection in emails without being constrained by emails or sentences structures [54,29,31]. Sentence based approaches: These approaches suppose that one activity is expressed at a sentence level. ...
... • Using a clustering algorithm in the vector space formed by the |Y | probabilities as in (Lemaire, Clérot, and Creff, 2015;Lemaire, Ismaili, et al., 2020;Zhdanov, 2019). ...
![[object Object]](https://i1.rgstatic.net/ii/profile.image/272728607424522-1442034975294_Q64/Vincent-Lemaire-3.jpg)
![[object Object]](https://i1.rgstatic.net/ii/profile.image/273638939164681-1442252015235_Q64/Antoine-Cornuejols.jpg)
![[object Object]](https://i1.rgstatic.net/ii/profile.image/272447152848899-1441967871299_Q64/Dominique-Gay.jpg)
... Traditionally, process mining always focuses on workflow discovery, that is, discovering tasks and the execution patterns between them from structured event logs (Y Liu et al., 2020). Traditional process mining provides strong support for business management; however, given the unstructured process data such as emails, meeting minutes, and conversation records, traditional process mining methodologies could not be applied directly (Elleuch M et al., 2020a). Recently, there have been several researchers tending to unstructured data. ...
... L'algorithme des k-moyennes est l'un des algorithmes de clustering le plus répandu dans la littérature [97], couramment utilisé pour la classification des signaux de DP [98]. Cet algorithme possède deux paramètres d'entrée qui sont le nombre de clusters à former ainsi que la matrice des données. ...
... One is to allow the Economy--litE and Economy-approaches, which use the confidence level of a binary classifier, to solve multi-classes problems. A second one is to use a supervised clustering technique to compute groups of times series (see Lemaire et al. 2020) in the Economy-K and Economy-multi-K approaches. Finally, we are working on the adaptation of these methods to the on-line detection of anomalies in a data stream. ...
... This definition differs from the more common view of 'supervised clustering' as a modification of standard clustering algorithms to identify class-uniform clusters, i.e., a supervised model selection of unsupervised clustering (e.g. Finley and Joachims, 2005;Michel et al., 2012;Ismaili et al., 2016). ...
... Utilizando uma abordagem supervisionada, pode-se avaliar a contribuição das variáveis presentes em cada um dos conjuntos de dados na formação dos agrupamento (Ismaili et al., 2014). Dessa forma, é possível compreender um agrupamento de dados de alta dimensão, o que é relevante ao analisar o conjunto Principais Dimensões que possui 85 dimensões. ...
