Algorithme pour la génération de documents personnalisés structurés Optimisation par fusion des classes d'utilisateurs

Abstract : The New Technologies of Information and Communication – NTIC – encourage the mass mailing of documents (to several thousands of people). Advertisers know that sending the same message to too many people at the same time increases the risk of the message becoming impersonal, and thus losing its interest and impact. It would therefore seem beneficial to personalize the message in order to increase its appeal and relevance. This personalization can be achieved by using various criteria to characterize the recipients. Based on these profiles, documents that are more appropriate to each kind of profile can be automatically generated. However if the criteria are too specific there is a risk of producing almost as many different documents as there are recipients. In this paper we propose a way to use the structure of the document to avoid this problem. Two users may be different with respect to a criterion C, but if this criterion is only relevant to one subpart of the document the system can produce the same document for both users except for this subpart, which will be specific for each of them. If users are grouped into equivalence classes for each part of the document, the system can then generate one subpart for each class instead of one subpart per user. In this paper we present an efficient algorithm for defining user classes for one part of a document, using only the criteria that are relevant to that part.
Document type :
Conference papers
Complete list of metadatas

Cited literature [3 references]  Display  Hide  Download

https://edutice.archives-ouvertes.fr/edutice-00000382
Contributor : Florence Galand <>
Submitted on : Wednesday, March 3, 2004 - 12:29:53 PM
Last modification on : Tuesday, September 17, 2019 - 9:35:59 AM
Long-term archiving on : Monday, March 29, 2010 - 5:59:08 PM

Identifiers

  • HAL Id : edutice-00000382, version 1

Collections

Citation

Vincent Ranwez, Sylvie Ranwez. Algorithme pour la génération de documents personnalisés structurés Optimisation par fusion des classes d'utilisateurs. NîmesTIC' 2001, Relation homme-système : complexe ?, Dec 2001, Nîmes, France. ⟨edutice-00000382⟩

Share

Metrics

Record views

371

Files downloads

257