An examination of natural language processing in Prolog for those who know Prolog but not linguistics, this book enables students to move quickly into writing and working in useful software. It features many working computer programs that implement subsystems of a natural language processor. These programs are designed to be understood in isolation from one another and are compatible with an Edinburgh-compatible Prolog implementation, such as Quintus, ESL, Arity and ALS.
It's a textbook on NLP from the early 90's (I'm a nerd), and I just got ahold of this book in 2022. For many topics, being so dated is a bad thing, but for NLP (in my opinion) it's good, since this book focuses on the structure & meaning of language. It predates the trendy neural network stuff. NLP with Prolog requires a deep understanding of the language(s) and a lot of current work tries to avoid this by making word salad and feeding it into the neural net black box.
As a bonus, the appendix is the best summary of the features of the Prolog language I've seen (though that appendix is pretty dense)