| Date: | 2021, November 30 |
| Time: | 10:00 a. m. |
| Place: | Online |
| Author: | Razzaghian, Negar |
| Title: | Translation of Natural Language Competency Questions into SPARQL-OWL Queries Using Neural Language Models |
As the formulation of competency questions from formal language needs special knowledge and familiarity with the query language such as SPARQL-OWL queries, domain experts prefer to formulate the competency questions from natural language instead of formal language. Thus, a neural translation machine for translation from a specific domain to another one is necessary. This work uses seq2seq as a recurrent neural network-based model (RNN) to translate competency questions into corresponding SPARQL-OWL queries. The encoder reads a sequence of words as a vector and summarizes it, and the decoder generates the output sequence, that is, SPARQL-OWL queries.