Mathematica-based natural language processing in applied mechanics

The probably novel possibility of using the computer algebra system Mathematica in NLP (natural language processing) is briefly studied in few elementary applications from applied mechanics. At first, Buchberger’s Mathematica-based Theorema system was found able to lead to a completely natural proof...

Full description

Bibliographic Details
Main Author: Ioakimidis, Nikolaos
Other Authors: Ιωακειμίδης, Νικόλαος
Format: Conference (paper)
Language:English
Published: 2018
Subjects:
Online Access:http://www.des.upatras.gr/amm/ioakimidis/personal/papers/ang1-p094--fullpaper.pdf
http://hdl.handle.net/10889/10851
Description
Summary:The probably novel possibility of using the computer algebra system Mathematica in NLP (natural language processing) is briefly studied in few elementary applications from applied mechanics. At first, Buchberger’s Mathematica-based Theorema system was found able to lead to a completely natural proof of an elementary propositional-type example from applied mechanics. The same system, Theorema, was also found useful for the syntactic analysis of related very simple sentences, a classical field of application of the Prolog logic-oriented language. The next step has been to use Prolog itself as an external package to Mathematica. Finally, elementary NLP has been achieved directly with Mathematica. The present results show that NLP is within Mathematica’s reach, offering the further advantage of performing symbolic computations (contrary to logic-oriented languages). Therefore, probably, gradually, Mathematica (independently of Theorema and Prolog) could lead to an automated environment of both text and formula checking (including symbols and computations) in applied mechanics and mechanical engineering in general.