# Automated Creation and Human-assisted Curation of Computable Scientific Models from Code and Text

arXiv: https://arxiv.org/abs/2202.13739 · HTML: https://mlsanity.com/paper/2202.13739/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 37 references that are in the map.
- First posted on arXiv 28 Jan 2022.

## Builds on (most cited first, 8 of 8)

- [BioBERT: a pre-trained biomedical language representation model for biomedical text mining](https://mlsanity.com/paper/1901.08746/) (2019)
- [SciBERT: A Pretrained Language Model for Scientific Text](https://mlsanity.com/paper/1903.10676/) (2019)
- [Neural Architectures for Named Entity Recognition](https://mlsanity.com/paper/1603.01360/) (2016)
- [Bidirectional LSTM-CRF Models for Sequence Tagging](https://mlsanity.com/paper/1508.01991/) (2015)
- [A Survey on Recent Advances in Named Entity Recognition from Deep Learning models](https://mlsanity.com/paper/1910.11470/) (2019)
- [Multi-Scale Attention with Dense Encoder for Handwritten Mathematical Expression Recognition](https://mlsanity.com/paper/1801.03530/) (2018)
- [SIRIUS-LTG-UiO at SemEval-2018 Task 7: Convolutional Neural Networks with Shortest Dependency Paths for Semantic Relation Extraction and Classification in Scientific Papers](https://mlsanity.com/paper/1804.08887/) (2018)
- [Unsupervised Construction of Knowledge Graphs From Text and Code](https://mlsanity.com/paper/1908.09354/) (2019)

## Abstract (arXiv)

Scientific models hold the key to better understanding and predicting the behavior of complex systems. The most comprehensive manifestation of a scientific model, including crucial assumptions and parameters that underpin its usability, is usually embedded in associated source code and documentation, which may employ a variety of (potentially outdated) programming practices and languages. Domain experts cannot gain a complete understanding of the implementation of a scientific model if they are not familiar with the code. Furthermore, rapid research and development iterations make it challenging to keep up with constantly evolving scientific model codebases. To address these challenges, we develop a system for the automated creation and human-assisted curation of a knowledge graph of computable scientific models that analyzes a model's code in the context of any associated inline comments and external documentation. Our system uses knowledge-driven as well as data-driven approaches to identify and extract relevant concepts from code and equations from textual documents to semantically annotate models using domain terminology. These models are converted into executable Python functions and then can further be composed into complex workflows to answer different forms of domain-driven questions. We present experimental results obtained using a dataset of code and associated text derived from NASA's Hypersonic Aerodynamics website.