ACL 2021 Social & Poster Session

1. "Must know before design and develop a chatbot in four domains: health, education, business and governance"
Hajar Zerouani*, Abdelhay Haqiq*^ and Bouchaib Bounabat*
*ALQUALSADI Team, Rabat IT Center, ENSIAS, Mohammed V University, Rabat, Morocco
^ITQAN Team, LyRICA Laboratory, ESI, Rabat, Morocco

Abstract: Chatbots are artificial intelligence systems that are designed to emulate the unstructured talks or ‘chats'. These dialogue systems communicate with users in natural language. They're now widely employed in a range of businesses or digital assistants to react to users' questions and requests via chat interfaces. Before we attempt to design a conversational AI that can interact with humans, we must first understand chatbots architecture, knowing the most important chatbot dialogue system in the history and which technologies used to develop it .This paper consists in order an overview of the published research work on chatbots, starting from history and chatbots architecture, discussing the different models and deep neural network algorithms used to improve natural language process and sentiment analysis for the chatbots, and classifying many pertained models and types of available datasets. It presents four domains where chatbots are useful:Health, Education, Businesses and Governance. It terminates by listing those technology’s limitations shortening its lifespan and its potential improvement for best performance.

2. "Knowledge-driven NLP: A New Path for Arabic NLP"
Houcemeddine Turki
Data Engineering and Semantics Research Unit, University of Sfax, Tunisia

Abstract: Nowadays, Arabic NLP largely depend on applying statistical methods on large corpuses to develop useful applications. However, knowledge graphs, ontologies and taxonomies are currently available as valuable resources for Arabic dialects but not effectively used in practice. In this brief poster, I explain how knowledge graphs can be useful for Arabic NLP and invite researchers to choose this path for their future research output.

3. "A Linguistic Analysis of Rhetorical Questions Forms and Functions: Towards NLP Application"
Yomna El Hossary
Trinity College, Dublin, Ireland

Abstract: This study aims at analysing the Rhetorical Questions, linguistically from a computational pragmatic perspective through decomposing the lexical, syntactic and discourse cues and interpret them systematically to form an architecture valid for the natural language processing . This could collaborate futuristically in enhancing machine understanding for humans linguistic intentions.

4. "Arabic Data Sources: Big Science Project🌸"
Maraim Masoud
Insight Centre for Data Analytics, National University of Ireland, Galway