A Scalable Database forSyntactic andLexical Resources

Paper in proceeding, 2025

We present a novel database engine tailored specifically for storing syntactic rules as well as lexical entries for natural language generation. The engine supports an unlimited number of languages which are moreover linked via an abstract syntax in a manner compatible with the Grammatical Framework. This makes it easier to generate content in several languages simultaneously. Dynamic addition and removal of data is possible. Transactions ensure the data integrity and enforce single writer – multiple readers policy. The writer does not block the readers.