This paper proposes an idea for developing a computational model of creative processes in design. This model facilitates and accelerates idea generation in the inventive design, increasing the solution space definition by suggesting technical actions and graphical triggers. The problem solver has to state the required design objective using any verbal action, then an automatic system generates an appropriate set of triggering actions indicating different ways of accomplishing that goal. In addition, for each verb is associated a list of evocative images indicating how that action can be implemented in space/time and through specific physical effects. The system is capable of handling the huge number of verbs that the English language offers. To select all functional verbs of the technical lexicon, the patent database has been processed using the most advanced text mining techniques. Among them, a customized version of Word2Vec model has been exploited to learn word/actions associations from a large corpus of patents. The article explains how the libraries have been created, the progress the software prototype and the results of a first validation campaign.

(2022). AI Based Patent Analyzer for Suggesting Solutive Actions and Graphical Triggers During Problem Solving . Retrieved from http://hdl.handle.net/10446/229069

AI Based Patent Analyzer for Suggesting Solutive Actions and Graphical Triggers During Problem Solving

Russo, Davide;
2022

Abstract

This paper proposes an idea for developing a computational model of creative processes in design. This model facilitates and accelerates idea generation in the inventive design, increasing the solution space definition by suggesting technical actions and graphical triggers. The problem solver has to state the required design objective using any verbal action, then an automatic system generates an appropriate set of triggering actions indicating different ways of accomplishing that goal. In addition, for each verb is associated a list of evocative images indicating how that action can be implemented in space/time and through specific physical effects. The system is capable of handling the huge number of verbs that the English language offers. To select all functional verbs of the technical lexicon, the patent database has been processed using the most advanced text mining techniques. Among them, a customized version of Word2Vec model has been exploited to learn word/actions associations from a large corpus of patents. The article explains how the libraries have been created, the progress the software prototype and the results of a first validation campaign.
File allegato/i alla scheda:
File Dimensione del file Formato  
978-3-031-17288-5_17.pdf

Solo gestori di archivio

Versione: publisher's version - versione editoriale
Licenza: Licenza default Aisberg
Dimensione del file 838.79 kB
Formato Adobe PDF
838.79 kB Adobe PDF   Visualizza/Apri
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

Aisberg ©2008 Servizi bibliotecari, Università degli studi di Bergamo | Terms of use/Condizioni di utilizzo

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10446/229069
Citazioni
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact