Fasilkom UI student team builds artificial intelligence to test legislation compliance

TIME.CO, Jakarta – Student group of the Faculty of Computer Science of the University of Indonesia (Fasilkom) (user interface) won first place in the Ria Statistics Competition and Data Science Festival (Satria Data) 2023 organized by the National Achievement Center, Ministry of Education, Culture, Research and Technology of the Republic of Indonesia in Malang on August 20-24 2023.

Reporting from the UI website, the Fasilkom UI student team named Three Neuron V2 consists of Bryan Tjandra, Nyoo Steven Christopher Handoko and Oey Joshua Jodrian. The victory obtained in the Big Data Challenge category was achieved thanks to the innovation of Team Three Neuron V2 in the creation of artificial intelligence (AI) capable of automatically testing the alignment between laws and regulations.

Legislation has a central role in regulating the rights and duties of every citizen. However, sometimes there is disharmony between laws and regulations, for example conflicting content of regulations which can threaten legal certainty for the community.

Three Neuron V2 team innovation can help legislators and policy makers keep these laws and regulations in harmony and not in conflict with each other.

This proposed AI-assisted solution leverages Legal Textual Entailment (LTE), a subfield method in natural language processing (NLP) and artificial intelligence that focuses on determining a logical relationship between two legal texts.

This solution is implemented using the LawID dataset, also created by the Three Neuron V2 team for LTE purposes. The dataset consists of tens of thousands of data items which includes primary data such as laws and regulations, starting at the level of the constitution, laws, derivative regulations, such as government regulations and ministerial regulations, and includes secondary data such as court decisions. Constitution.


In the elaboration and synthesis of regulatory documents, Team Three Neuron V2 uses different linguistic models to produce subtopics using the BERTopic technique. One of the models proposed in the experiment, namely BigBird-NLI, was applied to the primary LTE section of the LawID dataset and managed to achieve 99% accuracy in the validation data after going through a fine-tuning process using 800 pairs of document paragraphs. These verse pairs are labeled using semi-supervised labeling and large language models (LLMs).

The process found that most of the pairs of sentences in the regulatory document show consistency. While the percentage of inconsistency amounts to only about 0.6% of all pairs of sentences. Incongruent sentence pairs are usually semantically inconsistent without the need for deep context.

Fasilkom UI Principal Petrus Mursanto appreciated the students’ achievements in the Satria Data 2023 event. Meanwhile, the success of the team cannot be separated from the support of Panca supervisor Hadi Putra who is a legal tech expert of Fasilkom UI .

The Satria Data 2023 competition was attended by 600 students made up of 209 teams from 50 Indonesian universities. Some of them are IPB University, Ten November Institute of Technology, Bandung Institute of Technology, Gadjah Mada University, Airlangga University, Diponegoro University, Brawijaya University and others.

Editors’ Choice: Chevening Scholarship registration opens September 12, check requirements

Quoted From Many Source

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button