Hanna Abi Akl
Vice-Dean & Chief Academic Officer at Data ScienceTech Institute
Researcher with the INRIA Wimmics team
Data ScienceTech Institute
4 Rue de la Collégiale
75005 Paris, France
Projects, Software & Data
Projects
- KGstat, a project that combines querying knowledge graphs using SPARQL and applying statistical analysis and modeling using the R programming language
- FuelMi, the pet that takes care of you
- Data Consult Transforms Service Delivery and Customer Experience with Oracle, a project that automates incident detection and management by integrating Oracle Service Cloud
- Lebanon Smart City Experience, a project that launched the first smart city in Lebanon at BeitMisk
Software
- CLGC, a Python package to categorize syllogisms and translate first-order logical statements into different knowledge representation languages
- LOGIC-NL-CLIF-Flan-T5-Large, a Logic Language Model fine-tuned to verify syllogisms in natural language and CLIF notation
- LOGIC-NL-Flan-T5-Large, a Logic Language Model fine-tuned to verify syllogisms in natural language
- SHADOW, a LM-probing model that matches the best SPARQL query to fill missing object entities in a (subject entity, relation, object entity) triple over the Wikidata knowledge graph
- FLAN-T5 small-WordNet, a generative model that classifies terms into types in conjuction with Wikidata semantic information
- FLAN-T5 small-GeoNames, a generative model that classifies terms into geographical locations in conjuction with Wikidata semantic information
- PSYCHIC, an extractive question-answering neuro-symbolic model that answers a knowledge graph question with a SPARQL query
- FinISH, a model that classifies financial terms from domain ontology classes
- FReE, a model that classifies the type of relations in financial sentences
- FNP T5 D2T complete, a complete T5-base data to text model specialized for Finance NLG
- FNP T5 D2T simple, a simple T5-base data to text model specialized for Finance NLG
- Text smoothing, a generative model for financial text rewriting
- Code of other experiments on GitHub
Data
- P-FOLIO-KR, data set of syllogisms expressed in different knowledge representation languages
- FOLIO-KR, data set of syllogisms expressed in different knowledge representation languages
- WordNet Semantic Primes, data set of vectorized representations of knowledge graph information for the WordNet semantic tower
- GeoNames Semantic Primes, data set of vectorized representations of knowledge graph information for the GeoNames semantic tower
- FNP 2022 FinCausal, data set for the FinCausal 2022 shared task
- FNP 2021 FinCausal, data set for the FinCausal 2021 shared task
- FNP 2020 FinCausal, data set for the FinCausal 2020 shared task