Scholarlink

The use of Natural Language Generation (NLG) for various applications has been gaining momentum in recent years. NLG is also being used and developed for various applications in the education sector. As part of the PAGEL research project, an NLG tool will be implemented to provide answers and assignments in a web-based simulation for psychology students. The simulation is intended to make the effect of various configurable parameters on work design visible and tangible. To this end, the answers and tasks of fictitious chat partners should be designed to be as human as possible. This paper presents the background of the project and the specific system to be developed for the NLG algorithm. For this purpose, state of the art for templated-based NLG algorithms, conceptualization, implementation, and evaluation are presented in this paper for our research. The paper concludes with a summary.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Implementation of a Templated-Based NLG Algorithm and Configurator-Tool to Provide Assignments Based on Different Psychological Parameters for a Web-Based Simulation for Psychology Courses at a German University

  • 章节

  • 发表日期：

    2025年

查看全文
  • Ramona Srbecky,
  • Simon-Alexander Wetzel,
  • Filipe Ferreira,
  • Wieland Fraas,
  • Jan Dettmers,
  • Matthias Hemmje

摘要

The use of Natural Language Generation (NLG) for various applications has been gaining momentum in recent years. NLG is also being used and developed for various applications in the education sector. As part of the PAGEL research project, an NLG tool will be implemented to provide answers and assignments in a web-based simulation for psychology students. The simulation is intended to make the effect of various configurable parameters on work design visible and tangible. To this end, the answers and tasks of fictitious chat partners should be designed to be as human as possible. This paper presents the background of the project and the specific system to be developed for the NLG algorithm. For this purpose, state of the art for templated-based NLG algorithms, conceptualization, implementation, and evaluation are presented in this paper for our research. The paper concludes with a summary.