English. The paper describes our submissions to the task on PoS tagging for Italian Social Media Texts (PoSTWITA) at Evalita 2016. We compared two approaches: a traditional HMM trigram Pos tagger and a Deep Learning PoS tagger using both character-level and word-level embeddings. The character-level embeddings performed better proving that they can provide a finer representation of words that a...

A classic psycholinguistic problem is parole; in its modern formulation, performance . This paper looks at aspects of parole or performance from a new direction, arguing that it is organized as a dialectic of unlike cognitive modes, a dialectic of gesture imagery and language. This dialectic is the key to the evocation, organization, and ultimate execution of meaningful actions shaped to take l...

Traditional linguistic field research employs the costand time-consuming and possibly biased approach of gathering data by sending out researchers and filling in questionnaires. We improve linguistic field research using crowdsourcing techniques to remedy these issues and open research to a wider audience. On the online platform MetropolItalia, going public in August 2012, Italian native speake...

Le Chatbots sono applicazioni che simulano il dialogo intelligente con esseri umani. Il paradigma di realizzazione più utilizzato è l’approccio Stimolo-Risposta (S/R). Lo Stimolo è un insieme di frasi attese dalla chatbot, la Risposta sono le frasi restituite all’utente. Spesso tale approccio limita le potenzialità dell’interazione: la chatbot risponde solo quando la frase sottomessa dall’utent...

editoriale