Manuscrit<\/a><\/span><\/p>\nR\u00e9sum\u00e9<\/strong><\/span><\/p>\nCette th\u00e8se porte sur l\u2019impact de l\u2019usage de la s\u00e9mantique dans le processus de la classification supervis\u00e9e de textes. Cet impact est \u00e9valu\u00e9 au travers d\u2019une \u00e9tude exp\u00e9rimentale sur des documents issus du domaine m\u00e9dical et en utilisant UMLS (Unified Medical Language System) en tant que ressource s\u00e9mantique. Cette \u00e9valuation est faite selon quatre sc\u00e9narii exp\u00e9rimentaux d\u2019ajout de s\u00e9mantique \u00e0 plusieurs niveaux du processus de classification. Le premier sc\u00e9nario correspond \u00e0 la conceptualisation o\u00f9 le texte est enrichi avant indexation par des concepts correspondant dans UMLS ; le deuxi\u00e8me et le troisi\u00e8me sc\u00e9nario concernent l\u2019enrichissement des vecteurs repr\u00e9sentant les textes apr\u00e8s indexation dans un sac de concepts (BOC \u2013 bag of concepts) par des concepts similaires. Enfin le dernier sc\u00e9nario utilise la s\u00e9mantique au niveau de la pr\u00e9diction des classes, o\u00f9 les concepts ainsi que les relations entre eux, sont impliqu\u00e9s dans la prise de d\u00e9cision. Le premier sc\u00e9nario est test\u00e9 en utilisant trois des m\u00e9thodes de classification: Rocchio, NB et SVM. Les trois autres sc\u00e9narii sont uniquement test\u00e9s en utilisant Rocchio qui est le mieux \u00e0 m\u00eame d\u2019accueillir les modifications n\u00e9cessaires. Au travers de ces diff\u00e9rentes exp\u00e9rimentations nous avons tout d\u2019abord montr\u00e9 que des am\u00e9liorations significatives pouvaient \u00eatre obtenues avec la conceptualisation du texte avant l\u2019indexation. Ensuite, \u00e0 partir de repr\u00e9sentations vectorielles conceptualis\u00e9es, nous avons constat\u00e9 des am\u00e9liorations plus mod\u00e9r\u00e9es avec d\u2019une part l\u2019enrichissement s\u00e9mantique de cette repr\u00e9sentation vectorielle apr\u00e8s indexation, et d\u2019autre part l\u2019usage de mesures de similarit\u00e9 s\u00e9mantique en pr\u00e9diction.<\/span><\/p>\nAbstract<\/em><\/strong><\/span><\/p>\nThe main interest of this research is the effect of using semantics in the process of supervised text classification. This effect is evaluated through an experimental study on documents related to the medical domain using the UMLS (Unified Medical Language System) as a semantic resource. This evaluation follows four scenarios involving semantics at different steps of the classification process: the first scenario incorporates the conceptualization step where text is enriched with corresponding concepts from UMLS; both the second and the third scenarios concern enriching vectors that represent text as Bag of Concepts (BOC) with similar concepts; the last scenario considers using semantics during class prediction, where concepts as well as the relations between them are involved in decision making. We test the first scenario using three popular classification techniques: Rocchio, NB and SVM. We choose Rocchio for the other scenarios for its extendibility with semantics. According to experiment, results demonstrated significant improvement in classification performance using conceptualization before indexing. Moderate improvements are reported using conceptualized text representation with semantic enrichment after indexing or with semantic text-to-text semantic similarity measures for prediction.<\/span><\/em><\/span><\/p>\n![\"\"](\"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/wp-content\/uploads\/2021\/12\/Soutenance-Shereen-Albitar-300x225.jpeg\")
<\/p>\n
\n","protected":false},"excerpt":{"rendered":"
Shereen Albitar, \u00ab On the use of semantics in supervised text classification: application in the medical domain \u00bb. Th\u00e8se en informatique d’Aix-Marseille Universit\u00e9 soutenue le 12 d\u00e9c. 2013. Manuscrit R\u00e9sum\u00e9 Cette th\u00e8se porte sur l\u2019impact de l\u2019usage de la s\u00e9mantique dans le processus de la classification supervis\u00e9e de textes. Cet impact est \u00e9valu\u00e9 au travers … <\/p>\n
Continuer la lecture \u00ab\u00a0These Sheren Albitar\u00a0\u00bb<\/span><\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_crdt_document":"","footnotes":""},"class_list":["post-447","page","type-page","status-publish","hentry","entry"],"_links":{"self":[{"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/pages\/447","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/comments?post=447"}],"version-history":[{"count":16,"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/pages\/447\/revisions"}],"predecessor-version":[{"id":1199,"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/pages\/447\/revisions\/1199"}],"wp:attachment":[{"href":"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/index.php\/wp-json\/wp\/v2\/media?parent=447"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"\"](\"https:\/\/pageperso.lis-lab.fr\/bernard.espinasse\/wp-content\/uploads\/2021\/12\/Sheren-these.png\")
<\/p>\n