Home |  English version |  Mappa |  Commenti |  Sondaggio |  Staff |  Contattaci Cerca nel sito  
Istituto di scienza e tecnologie dell'informazione "Alessandro Faedo"

Torna all'elenco Report e working Paper anno 2019

Report e working Paper

Tipo: Rapporto di ricerca (Research report)

Titolo: Funnelling: a new ensemble method for heterogeneous transfer learning and its application to polylingual text classification

Anno di pubblicazione: 2019

Formato: Elettronico

Autori: Esuli A.; Moreo Fernandez A.D.; Sebastiani F.

Affiliazioni autori: CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy

Autori CNR:

  • ANDREA ESULI
  • ALEJANDRO DAVID MOREO FERNANDEZ
  • FABRIZIO SEBASTIANI

Lingua: inglese

Sintesi: Polylingual Text Classification (PLC) consists of automatically classifying, according to a common set C of classes, documents each written in one of a set of languages L, and doing so more accurately than when naively classifying each document via its corresponding language-specific classifier. In order to obtain an increase in the classification accuracy for a given language, the system thus needs to also leverage the training examples written in the other languages. We tackle multilabel PLC via funnelling, a new ensemble learning method that we propose here. Funnelling consists of generating a two-tier classification system where all documents, irrespectively of language, are classified by the same (2nd-tier) classifier. For this classifier all documents are represented in a common, language-independent feature space consisting of the posterior probabilities generated by 1st-tier, language-dependent classifiers. This allows the classification of all test documents, of any language, to benefit from the information present in all training documents, of any language. We present substantial experiments, run on publicly available polylingual text collections, in which funnelling is shown to significantly outperform a number of state-of-the-art baselines. All code and datasets (in vector form) are made publicly available.

Lingua sintesi: eng

Pagine totali: 28

Parole chiave:

  • Ensemble learning; Heterogeneous transfer learning

URL: https://arxiv.org/abs/1901.11459

Strutture CNR:

Moduli:

Allegati: Funnelling: a new ensemble method for heterogeneous transfer learning ... (application/pdf)

 
Torna indietro Richiedi modifiche Invia per email Stampa
Home Il CNR  |  I servizi News |   Eventi | Istituti |  Focus