We present the system developed at FBK for the EVALITA 2016 Shared Task “QA4FAQ – Question Answering for Frequently Asked Questions”. A peculiar characteristic of this task is the total absence of training data, so we created a meaningful representation of the data using only word embeddings. We present the system as well as the results of the two submitted runs, and a qualitative analysis of them.
Presentiamo il sistema sviluppato presso FBK per la risoluzione del task EVALITA 2016 “QA4FAQ - Question Answering for Frequently Asked Questions”. Una caratteristica peculiare di questo task ´e la totale mancanza di dati di training, pertanto abbiamo creato una rappresentazione significativa dei dati utilizzando solamente word embeddings. Presentiamo il sistema assieme ai risultati ottenuti dalle due esecuzioni che abbiamo inviato e un’analisi qualitativa dei risultati stessi.
Tweaking Word Embeddings for FAQ Ranking
Rocha Fonseca, Erick;Magnolini, Simone;Feltracco, Anna;Qwaider, Mohammed R H;Magnini, Bernardo
2016-01-01
Abstract
We present the system developed at FBK for the EVALITA 2016 Shared Task “QA4FAQ – Question Answering for Frequently Asked Questions”. A peculiar characteristic of this task is the total absence of training data, so we created a meaningful representation of the data using only word embeddings. We present the system as well as the results of the two submitted runs, and a qualitative analysis of them.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
