Using Grammar Masking to Ensure Syntactic Validity in LLM-based Modeling Tasks

Low-code development platforms (LCDPs) are becoming increasingly important in industry, which confronts us in academic teaching with the challenge of educating students in the basic principles, critical engagement, and evaluation of LCDPs. This leads us to the question, how to teach the usage of different LCDPs during an university course. The short time frame of university-level courses makes it challenging to teach more than only one LCDP. In our teaching approach, students use two different LCDPs and create a web-application with both of them. Firstly, we require the students to define a target application with common modeling languages, next they use the first LCDP, at about half the time they switch to the second LCDP and present their findings of the differences in methodology and development processes at the end. We discuss this approach, show survey results from the participants, and explain lessons learned. This concept allows students critical engagement with LCDPs and model-driven software engineering. Supervisors get an insight into the learnability of each LCDP and how novices adapt to different domain-specific languages and their notations.

Authors

Institutions

Publication Details

Published
2024-09-22
DOI
https://doi.org/10.1145/3652620.3687805
Citations
9
Primary Topic
Natural Language Processing Techniques
Type
article
Field-Weighted Citation Impact
3.13
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Using Grammar Masking to Ensure Syntactic Validity in LLM-based Modeling Tasks

Bernhard Rumpe⋆, Jan Reimer, Lukas Netz
9 citations
Natural Language Processing Techniques
3.13
article

Using Grammar Masking to Ensure Syntactic Validity in LLM-based Modeling Tasks

Bernhard Rumpe⋆, Jan Reimer, Lukas Netz
article en
9 citations

Abstract

Low-code development platforms (LCDPs) are becoming increasingly important in industry, which confronts us in academic teaching with the challenge of educating students in the basic principles, critical engagement, and evaluation of LCDPs. This leads us to the question, how to teach the usage of different LCDPs during an university course. The short time frame of university-level courses makes it challenging to teach more than only one LCDP. In our teaching approach, students use two different LCDPs and create a web-application with both of them. Firstly, we require the students to define a target application with common modeling languages, next they use the first LCDP, at about half the time they switch to the second LCDP and present their findings of the differences in methodology and development processes at the end. We discuss this approach, show survey results from the participants, and explain lessons learned. This concept allows students critical engagement with LCDPs and model-driven software engineering. Supervisors get an insight into the learnability of each LCDP and how novices adapt to different domain-specific languages and their notations.

RWTH Aachen University (DE)
Openalex Percentile: Top 7%
Natural Language Processing Techniques
3.13
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Using Grammar Masking to Ensure Syntactic Validity in LLM-based Modeling Tasks — Bernhard Rumpe⋆, Jan Reimer, et al. · (2024) | TGRS Research Map | TGRS