Publications

Here you can find a list of my publications. Please feel free to check out their .pdf versions.

National journal papers

  • NEW! Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Prédire l'intensité de contradiction dans les commentaires : faible, forte ou très forte ?
    In: Revue d'Intelligence Artificielle (RIA 2020).
    (to appear)
  • NEW! Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Prédire l'intensité de contradiction dans les commentaires : faible, forte ou très forte ?
    In: Le Bulletin n°101 Association Française pour l'Intelligence Artificielle (AFIA 2019).
    Access:

International conference papers

  • NEW! Ismail Badache.
    Exploring Differences in the Impact of Users' Traces on Arabic and English Facebook Search.
    In: IEEE/WIC/ACM International Conference on Web Intelligence (WI 2019).
    (to appear)
  • NEW! Ismail Badache.
    Users' Traces for Enhancing Arabic Facebook Search.
    In: ACM Hypertext and Social Media (HT 2019).
    (to appear)
  • Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Predicting Contradiction Intensity: Low, Strong or Very Strong?.
    In: ACM SIGIR Special Interest Group on Information Retrieval (SIGIR 2018).
    Access:
  • Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Contradiction in Reviews: is it Strong or Low?.
    In: BroDyn Workshop at European Conference on Information Retrieval (BroDyn@ECIR 2018).
    Access:
  • Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Finding and Quantifying Temporal-Aware Contradiction in Reviews.
    In: Asia Information Retrieval Societies (AIRS 2017).
    Access:
  • Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Harnessing Ratings and Aspect-Sentiment to Estimate Contradiction Intensity in Temporal-Related Reviews. In: Conference on Knowledge-Based and Intelligent Information and Engineering Systems (KES 2017).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Emotional Social Signals for Search Ranking
    In: ACM SIGIR Special Interest Group on Information Retrieval (SIGIR 2017).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Fresh and Diverse Social Signals: Any Impacts on Search?
    In: ACM Conference on Human Information Interaction and Retrieval (CHIIR 2017).
    Access:
  • Mélanie Imhof, Ismail Badache, Mohand Boughanem.
    Multimodal Social Book Search.
    In: Conference on Multilingual and Multimodal Information Access Evaluation (CLEF 2015).
    Access:
  • Ismail Badache, Mohand Boughanem.
    A Priori Relevance Based On Quality and Diversity Of Social Signals.
    In: ACM SIGIR Special Interest Group on Information Retrieval (SIGIR 2015).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Document Priors Based On Time-Sensitive Social Signals.
    In: European Conference on Information Retrieval (ECIR 2015).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Social Priors to Estimate Relevance of a Resource.
    In: ACM Information Interaction in context (IIiX 2014).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Harnessing Social Signals to Enhance a Search.
    In: IEEE/WIC/ACM International Conference on Web Intelligence (WIC 2014).
    Access:

National conference papers

  • NEW! Ismail Badache, Aya Abu-Thaher, Mariam Hamdan, Lara Abu-Jaish.
    Recherche d’Information Sociale en Langue Arabe : Cas de Facebook
    In: COnférence en Recherche d'Information et Applications (CORIA 2019).
    Access:
  • Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Prédire l'intensité de contradiction dans les commentaires : faible, forte ou très forte ?
    In: journées francophones d'Ingénierie des Connaissances (IC 2018) - 2nd Best Paper.
    Access:
  • Ismail Badache, Sébastien Fournier, Adrian Chifu.
    Détection de contradiction dans les commentaires.
    In: COnférence en Recherche d'Information et Applications (CORIA 2017).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Les Signaux Sociaux Émotionnels : Quel impact sur la recherche d'information ?.
    In: COnférence en Recherche d'Information et Applications (CORIA 2017).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Pertinence a Priori Basée sur la Diversité et la Temporalité des Signaux Sociaux.
    In: COnférence en Recherche d'Information et Applications (CORIA 2015).
    Access:
  • Ismail Badache, Mohand Boughanem.
    Exploitation de signaux sociaux pour estimer la pertinence a priori d'une ressource.
    In: COnférence en Recherche d'Information et Applications (CORIA-CIFED 2014).
    Access:
  • Ismail Badache, Mohand Boughanem.
    RI sociale: intégration de propriétés sociales dans un modèle de recherche.
    In: COnférence en Recherche d'Information et Applications (CORIA 2013).
    Access:

Other publications

  • Ismail Badache.
    Intensité de contradiction dans les commentaires.
    Seminary (EHESS School of Advanced Studies in the Social Sciences - Paris). Avril 2018.
  • Access:
  • Ismail Badache.
    Les contenus sociaux : Quel impact sur le processus de RI et la quantification de contradiction.
    Seminary (Team DRIM, LIRIS - Lyon). Mars 2018.
  • Access:
  • Ismail Badache.
    Exploitation des contenus sociaux dans des tâches de RI et de détection de contradiction.
    Seminary (Team TALEP, LIF - Marseille). April 2017.
  • Ismail Badache.
    Social Signals: Any Impacts in Search?.
    Conference (Journée de la Science des Données, IRIT - Toulouse). April 2016.
  • Ismail Badache.
    RI Sociale : Exploitation des signaux sociaux pour améliorer la recherche d'information.
    Seminary (DocToMe, Toulouse). June 2014.

Slideshare

Here you can find a list of my conferences/seminars presentations. Consult the full list on Slides

Teachings

Here is some of my teaching courses (French).


Système intégré de gestion de bibliothèque


Techniques d'accès à l'information : les SRI


Démarches et outils bibliographiques : Zotero 5.0


E-Portfolio : Votre vitrine numérique


Numérique et éducation aux médias


Systèmes d'information de gestion (PGI, ERP en anglais)


Systèmes de gestion de bases de données


Analyse et conception des systèmes d'information


Programmation Web (HTML,PHP, J2EE)


Recherche d'information et réseaux sociaux


Environnement numérique (Pix, C2i, MS-Office)


Accompagnement et initiation aux outils informatique


Tools

Here you can find a list of my some online softwares.

  • Arabic Sentiment Analysis
    Arabic Sentiment Analysis
  • 1) Arabic Sentiment Analysis (تحليل المشاعر الكتابية باللغة العربية).
    This work has just been born ... the best will come.


  • Emotion Analysis
    Emotion Analysis
  • 2) Emotion Analysis (Happy, Sad, Angry, Excited, Sarcasm or Fear).
    Please pass short texts (tweets, reviews, Posts, etc.)


  • Sentiment Analysis
    Sentiment Analysis
  • 3) Sentiment model based on Fast and accurate Naive Bayes classification.
    This tool works by examining individual words and short sequences of words (n-grams) and comparing them with a probability model. The probability model is built on a set of pre-labeled tests of IMDb movie reviews. It can also detect negations in sentences, ie the phrase "not bad" will be classified as positive despite having two individual words with a negative feeling. Some adverbs like "very", "absolutely" are taken into account in sentiment estimation, as well as some implicitly negative phrases like "I regret ..." can be identified as sequences carrying a negative polarity.


  • Social Count
    Social Count
  • 4) Social data extractor from several social networks by using URL.

Projects

Here you can find a few details of my research projects.

(1) The ADNVideo Project

  • Description: The ADNVideo Project aims to develop multimodal tools for video analysis and recommendation. It focuses on jointly processing audio, speech transcripts, images, scenes, text overlays and user feedback. Using as starting point the corpus, annotations and approaches developed during the REPERE challenge, this project aims at going beyond indexing at single modalities by incorporating information retrieval methods, not only from broadcast television shows, but more generally on video documents requiring multimodal scene analysis. The novelty of our project is to combine and correlate information from different sources to enhance the description of the content.
  • Company: Kalyzee
  • Research Labs: LIS and LIF (Aix-Marseille University)
  • Contributors: Ismaïl BADACHE, Sébastien FOURNIER and Adrian-Gabriel CHIFU
  • Deadline: December 31th, 2017
  • Funding: The ADNVIDEO is funded in the framework of A*MIDEX.
  • Our Tasks: Contradictions Detection, Sentiments Analysis and Aspects Detection.
  • Work Overview (Datasets, Stats, Results, etc): Click Here

(2) Social IR : Exploitation of social signals to enhance IR

  • Description: My research work focuses on social information retrievial (SIR). In this context, I am interested in ranking models to enhance IR in social networks and the Web. The problem of SIR that we want to solve during this project is to answer several questions:
  • 1) The first concerns the development of tools to meet users needs:
  • - What are the information needs of social media users?
  • - What models of social information retrieval?
  • 2) The second concerns the identification and exploitation of social information in IR tasks :
  • - Can these social signals or UGC help the underlying search systems for guiding their users to reach a better quality or more relevant content?
  • - How can we convert these UGC into social properties?
  • - What are the useful social properties to be taken into account to improve the resources search (e.g., Web pages, videos)?
  • - What is the impact of both emotionnel signals and comments sentiments on search ranking?
  • So, our contrubutions aim to estimate relevance of information through the qualitative, temporal and relationship properties. Thus, the analysis and comparison of communities and their properties and behaviors in a model of the type ≺ user, resource, action, time ≻
  • Research Labs: LIS, IRIT and ZHAW
  • Contributors: Ismaïl BADACHE, Mohand BOUGHANEM and Mélanie IMHOF
  • Deadline: December 31th, 2018
  • Our Tasks: Social Information Retrieval, Sentiments Analysis and Opinion Detection.
  • INEX IMDb Dataset (Archive .gz): Save as
  • INEX IMDb Dataset (Solr Format): Save as
  • INEX IMDb Topics: Save as
  • INEX IMDb Results (30 Topics - with social signals): Save as
  • INEX SBS Dataset: Click here

News

Here you can find a some news about computer science.

"Fake news ou infox" : lequel va s'imposer ?

  • Il y a une expression américaine qui est fort à la mode dans la presse française depuis la campagne présidentielle de Donald Trump, il s’agit de : "fake news". Pour tenter de couper l’herbe sous le pied à cet anglicisme, "qui désigne un ensemble de procédés contribuant à la désinformation du public", la Commission d’enrichissement de la langue française vient de proposer de le remplacer par un néologisme : "infox".
  • Le mot est assez bien trouvé selon moi. Reste à voir si la réaction a été suffisamment rapide pour que cette "infox", un mot-valise basé évidemment sur "info" et "intox", s’installe dans notre vocabulaire.
  • La Commission d’enrichissement de la langue est une émanation du ministère de la culture, créée en 1996. "Pour demeurer vivante, explique la Commission, une langue doit être en mesure d’exprimer le monde moderne dans toute sa diversité et sa complexité. Chaque année, dans notre monde désormais dominé par la technique, des milliers de notions et de réalités nouvelles apparaissent, qu’il faut pouvoir comprendre et nommer".
  • Ecrit Par : M. Gilbert
  • Publié le : 21/10/2018 à 09:35

SIGIR 2018: Turning Clicks into Purchases: Revenue Optimization for Product Search in E-Commerce

  • Abstract: In recent years, product search engines have emerged as a key factor for online businesses. According to a recent survey, over 55% of online customers begin their online shopping journey by searching on an E-Commerce (EC) website like Amazon as opposed to a generic web search engine like Google. Information retrieval research to date has been focused on optimizing search ranking algorithms for web documents while little attention has been paid to product search. There are several intrinsic differences between web search and product search that make the direct application of traditional search ranking algorithms to EC search platforms difficult. First, the success of web and product search is measured differently; one seeks to optimize for relevance while the other must optimize for both relevance and revenue. Second, when using real-world EC transaction data, there is no access to manually annotated labels. In this paper, we address these differences with a novel learning framework for EC product search called LETORIF (LEarning TO Rank with Implicit Feedback). In this framework, we utilize implicit user feedback signals (such as user clicks and purchases) and jointly model the different stages of the shopping journey to optimize for EC sales revenue. We conduct experiments on real-world EC transaction data and introduce a a new evaluation metric to estimate expected revenue after re-ranking. Experimental results show that LETORIF outperforms top competitors in improving purchase rates and total revenue earned.
  • Authors: Liang Wu, Diane Hu, Liangjie Hong and Huan Liu
  • Published: July 08th, 2018