Research Project
MARLENE: semantic analysis of texts to execute NLP
A tool developed by our Research and Developement Lab leveraging AI to enable multilingual text analysis and understanding.
Where: international
Where: international
Challenge
Natural language understanding represents one of the most important challenges for Artificial Intelligence. Starting from initial statistics based algorithms, the computational linguistics has evolved through semantically based approaches able to model domain ontologies.
Approach
The advent of the Big Data infrastructure has significantly boosted the Natural Language Processing (NLP) through Machine Learning and Deep Learning techniques. In this context, MARLENE improves the understanding of textual contents fusing together a semantic approach, based on extense Semantic Net and performing ML and AI algorithms to offer a large plethora of multilingual NLP services.
Digital Ecosystem
Solution
Digital Ecosystem
MARLENE provides a multilingual syntactic and textual semantic analysis framework to perform the entire NLP chain. The framework supplies NLP services for parsing, tagging, word sense disambiguation, categorization, clustering, summarization, text similarity, sentiment analysis, emotion recognition, textual pattern identification.
MARLENE leverages AI to enable text understanding on several languages including Italian, English, Spanish, Portuguese, Dutch, Arabic and French. MARLENE embeds a large SemanticNet of concepts, built integrating and extending both open and property semantic nets, to perform word sense disambiguation and term expansion and solving specific semantic issues.
Results
Large set of services for NLP
Leveraging Artificial Intelligence
Text understanding on several languages (EU and non-EU)
Impacts
Technologies
AI & Data
![]()
The power of big data leads services, products and processes to a higher level of "intelligence".
To know more
Recommended for you
Explore additional content associated with the topic
