The dynamic pointer to effect is an IR tool that searches for a scientific effect capable of providing the required function. At difference from static pointers which are entirely pre-built databases, the dynamic type is based on the integration of a pre-built effects library with a semantic search engine that scans technical and scientific literature in order to suggest documents describing relevant effects to realize a certain function. As result, the search engine can use any function in form of a pair verb-object, where the relations between functions and effects are always updated with the updating of the literature and the suggested list of effects is always tailored only on the specific function of the user, avoiding huge lists that make the tool unusable for low precision. The key aspect of the dynamic database is that recall and precision problems can be decoupled, in fact, recall depends on the completeness of effect library and the expansion of the query, while precision is affected by the IR algorithms of the search engine. The present article presents a comparison of the most representative pointers to effects (developed by Altshuller and later) together with an analysis of the open problems and main limitations of these tools shown on three case studies selected in order to bring out the main limitations of effects databases.
(2018). Tech-Finder: a Dynamic Pointer to Effects . In INNOVATOR. Retrieved from http://hdl.handle.net/10446/116287
Tech-Finder: a Dynamic Pointer to Effects
Russo, Davide;Montecchi, Tiziano;Caputi, Antonio
2018-01-01
Abstract
The dynamic pointer to effect is an IR tool that searches for a scientific effect capable of providing the required function. At difference from static pointers which are entirely pre-built databases, the dynamic type is based on the integration of a pre-built effects library with a semantic search engine that scans technical and scientific literature in order to suggest documents describing relevant effects to realize a certain function. As result, the search engine can use any function in form of a pair verb-object, where the relations between functions and effects are always updated with the updating of the literature and the suggested list of effects is always tailored only on the specific function of the user, avoiding huge lists that make the tool unusable for low precision. The key aspect of the dynamic database is that recall and precision problems can be decoupled, in fact, recall depends on the completeness of effect library and the expansion of the query, while precision is affected by the IR algorithms of the search engine. The present article presents a comparison of the most representative pointers to effects (developed by Altshuller and later) together with an analysis of the open problems and main limitations of these tools shown on three case studies selected in order to bring out the main limitations of effects databases.File | Dimensione del file | Formato | |
---|---|---|---|
selection.pdf
accesso aperto
Versione:
publisher's version - versione editoriale
Licenza:
Licenza default Aisberg
Dimensione del file
2.17 MB
Formato
Adobe PDF
|
2.17 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
Aisberg ©2008 Servizi bibliotecari, Università degli studi di Bergamo | Terms of use/Condizioni di utilizzo