This paper describes the framework proposed by the UNIMIB Team for the task of Named Entity Recognition and Linking of Italian Tweets (NEEL-IT). The proposed pipeline, which represents an entry level system, is composed of three main steps: (1) Named Entity Recognition using Conditional Random Fields, (2) Named Entity Linking by considering both Supervised and Neural-Network Language models, and (3) NIL clustering byusing a graph-based approach.
Questo articolo descrive il sistema proposto dal gruppo UNIMIB per il task di Named Entity Recognition and Linking applicato a tweet in lingua italiana (NEEL-IT). Il sistema, che rappresenta un approccio iniziale al problema, è costituito da tre passaggi fondamentali: (1) Named Entity Recognition tramite l’utilizzo di Conditional Random Fields, (2) Named Entity Linking considerando sia approcci supervisionati sia modelli di linguaggio basati su reti neurali, e (3) NIL clustering tramite un approccio basato su grafi.
Cecchini, F. M., Fersini, E., Manchanda, P., Messina, E., Nozza, D., Palmonari, M., Sas, C., UNIMIB@NEEL-IT: Named Entity Recognition and Linking of Italian Tweets, Comunicazione, in EVALITA Evaluation of NLP and SPeech Tools for Italian - Proceedings of the Final Workshop, (Naples, Italy, 07-07 December 2016), aAccademia University Press, Torino 2016:1749 54-59. 10.4000/books.aaccademia.1938 [http://hdl.handle.net/10807/122102]
UNIMIB@NEEL-IT: Named Entity Recognition and Linking of Italian Tweets
Cecchini, Flavio Massimiliano;
2016
Abstract
This paper describes the framework proposed by the UNIMIB Team for the task of Named Entity Recognition and Linking of Italian Tweets (NEEL-IT). The proposed pipeline, which represents an entry level system, is composed of three main steps: (1) Named Entity Recognition using Conditional Random Fields, (2) Named Entity Linking by considering both Supervised and Neural-Network Language models, and (3) NIL clustering byusing a graph-based approach.File | Dimensione | Formato | |
---|---|---|---|
3d52d8deb606fb432186e276ad6743e4f6ca.pdf
accesso aperto
Tipologia file ?:
Versione Editoriale (PDF)
Licenza:
Creative commons
Dimensione
211.26 kB
Formato
Adobe PDF
|
211.26 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.