indoler

This repository contains the dataset used in our undergraduate thesis. The dataset contains 993 annotated court decission document. The document was taken from Decision of the Supreme Court of Indonesia. the documents have also been tokenized and cleaned. The data is stored using JSONL format with fields as follows:

{
	"id": "the internal id of the document",
	"owner": "the person who annotate the document",
	"lawyer": "whether the defendant has a lawyer present",
	"verdict": "The type of verdict given in the document (guilty, bebas, or lepas)",
	"indictment": "The type of indictment presented to the court (Tunggal, subsider, komul, alternatif, kombinasi, or gabungan)",
	"text": "Word-splitted text",
	"text-tags": "IOB formatted label"
}

Here's two data from the dataset:

[
	{
		"id": 2743,
		"owner": "agree",
		"lawyer": False,
		"verdict": "guilty",
		"indictment": "subsider",
		"text-tags": ["O", "O", "B-Nomor Putusan", "I-Nomor Putusan", "O", "O", "O", "O", "O", "O", "O", "B-Nama Pengadilan", "I-Nama Pengadilan", "I-Nama Pengadilan", ...],
		"text": ["\ufeffputusan", "nomor", "325/pid.b/2015/pn", "bwi", "demi", "keadilan", "berdasarkan", "ketuhanan", "yang", "maha", "esa", "pengadilan", "negeri", "banyuwangi", ...]
	},
	{
		"id": 2744,
		"owner": "agree",
		"lawyer": False,
		"verdict": "guilty",
		"indictment": "alternatif",
		"text-tags": ["O", "O", "B-Nomor Putusan", "I-Nomor Putusan", "O", "O", "O", "O", "O", "O", "O", "B-Nama Pengadilan", "I-Nama Pengadilan", "I-Nama Pengadilan", ...],
		"text": ["\ufeffputusan", "nomor", "285/pid.b/2016/pn", "sda", "demi", "keadilan", "berdasarkan", "ketuhanan", "yang", "maha", "esa", "pengadilan", "negeri", "sidoarjo", ...]
	}
]

We also provide the data we used in each split in CSV format.

Contact

evi.y [at] cs.ui.ac.id

naradhipabhary27 [at] gmail.com