Researcher identifiers


Number of documents


Xavier Bost

Short biography. With a background in Humanities, Xavier Bost received his PhD from Avignon University in 2016: in the Computer Science Laboratory of Avignon (LIA), his work focused on speaker detection, multimedia modeling and movie summarization. Since 2017, he has been a research engineer in multimedia retrieval at Orkis SAS (Aix-en-Provence, France). Previously, he had been teaching philosophy in high school, and began taking interest in computer science when attempting to automate processes used in historical linguistics.

Research interests. His main research interests include multimedia retrieval, natural langage processing, and social network analysis.


  • Email:
  • Phone: +33 (0)4 42 61 23 20
  • Address: 610 rue Georges Claude, 13290 Aix en Provence, France



Journal articles2 documents

  • Vincent Labatut, Xavier Bost. Extraction and Analysis of Fictional Character Networks: A Survey. ACM Computing Surveys, Association for Computing Machinery, In press. ⟨hal-02173918⟩
  • X. Bost, G. Senay, M. El-Bèze, R. de Mori. Multiple topic identification in human/human conversations. Computer Speech and Language, Elsevier, 2015, 34 (1), pp.18-42. ⟨10.1016/j.csl.2015.03.006⟩. ⟨hal-01956840v2⟩

Book sections1 document

Conference papers10 documents

  • Titouan Parcollet, Mohamed Morchid, Xavier Bost, Georges Linarès. M2H-GAN: A GAN-based Mapping from Machine to Human Transcripts for Speech Understanding. INTERSPEECH 2019, Sep 2019, Gratz, Austria. ⟨hal-02107674v2⟩
  • Xavier Bost, Vincent Labatut, Serigne Gueye, Georges Linarès. Extraction de réseaux dynamiques conversationnels par lissage narratif. 7ème Conférence sur les modèles et l'analyse de réseaux : approches mathématiques et informatiques (MARAMI), Oct 2016, Cergy-Pontoise, France. ⟨hal-01385215⟩
  • Xavier Bost, Vincent Labatut, Serigne Gueye, Georges Linarès. Narrative Smoothing: Dynamic Conversational Network for the Analysis of TV Series Plots. DyNo: 2nd International Workshop on Dynamics in Networks, in conjunction with the 2016 IEEE/ACM International Conference ASONAM, Aug 2016, San Francisco, United States. pp.1111-1118, ⟨10.1109/ASONAM.2016.7752379⟩. ⟨hal-01276708v5⟩
  • Xavier Bost, Georges Linares. Détection de locuteurs dans les séries TV. Coria 2015, Mar 2015, Paris, France. ⟨hal-01313094⟩
  • Xavier Bost, Georges Linarès, Serigne Gueye. Audiovisual speaker diarization of TV series. 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Apr 2015, Brisbane, Australia. pp.4799-4803, ⟨10.1109/ICASSP.2015.7178882⟩. ⟨hal-01313080v2⟩
  • Xavier Bost, Marc El-Bèze, Renato de Mori. Catégorisation multi-thématique de dialogues téléphoniques. XXXe édition des Journées d’Études sur la Parole (JEP 2014), Jun 2014, Le Mans, France. ⟨hal-01967848⟩
  • Xavier Bost, Georges Linares. Constrained speaker diarization of TV series based on visual patterns. 2014 IEEE Spoken Language Technology Workshop (SLT), Dec 2014, South Lake Tahoe, United States. pp.390-395, ⟨10.1109/SLT.2014.7078606⟩. ⟨hal-01957900v2⟩
  • Jean Valère Cossu, Benjamin Bigot, Ludovic Bonnefoy, Mohamed Morchid, Xavier Bost, et al.. LIA@RepLab 2013. Replab: An evaluation campaign for Online Reputation Management Systems, Fourth International Conference of the CLEF initiative, Sep 2013, Valencia, Spain. ⟨hal-01967853⟩
  • Xavier Bost, Marc El Bèze, Renato de Mori. Multiple topic identification in telephone conversations. Interspeech, Aug 2013, Lyon, France. ⟨hal-01313070v2⟩
  • Xavier Bost, Ilaria Brunetti, Luis Adrian Cabrera Diego, Jean-Valère Cossu, Andréa Linhares, et al.. Systèmes du LIA à DEFT'13. DEFT2013, Jun 2013, Les Sables d'Olonne, France. ⟨hal-01313065⟩

Theses1 document

  • Xavier Bost. A storytelling machine?: Automatic video summarization: the case of TV series. . Computer Science [cs]. Université d'Avignon et des Pays de Vaucluse, 2016. English. ⟨tel-01402549⟩

Preprints, Working Papers, ...1 document

  • Jean-Valère Cossu, Ludovic Bonnefoy, Xavier Bost, Marc El Bèze. How to merge three different methods for information filtering ?. 2015. ⟨hal-01220188⟩