UPV/EHUko eZerbitzu baten modelatzea ikasketa automatikoaren bidez

  1. Yera, Ainhoa
  2. Perona, Iñigo
  3. Arbelaitz, Olatz
  4. Muguerza, Javier
Libro:
II. Ikergazte nazioarteko ikerketa euskaraz: Kongresuko artikulu-bilduma.
  1. Iñaki Alegria (ed. lit.)
  2. Ainhoa Latatu (ed. lit.)
  3. Miren Josu Omaetxebarria (ed. lit.)
  4. Patxi Salaberri (ed. lit.)

Editorial: Udako Euskal Unibertsitatea, UEU = Universidad Vasca de Verano

ISBN: 978-84-8438-627-8 978-84-8438-632-2

Año de publicación: 2017

Título del volumen: Ingenieritza eta Arkitektura

Tomo: 5

Volumen: 5

Páginas: 111-118

Congreso: Ikergazte. Nazioarteko Ikerketa Euskaraz (2. 2017. Iruñea)

Tipo: Aportación congreso

Resumen

In this work we have analyzed the enrollment eService navigation of the UPV/EHU and using data mining techniques we have attempted to automatically perform navigation sessions classification. The results show that we are able to detect the defined success and failure navigation behaviours. For example, more than 90 % of the sessions of the clusters labelled as success are of success type and in the failure case, around 90%. Besides, using supervised learning we are able to automatically distinguish the two nabigation types with an accuracy rate of 96 %. Thus, we think that this research is a suitable basis to improve the eService analyzed in a near future.