機器學習於中文法律文件之標記與分類

Abstract

人工智慧在於法學領域所發展出的法學資訊系統在於提供參考資訊以協助司法審判，重要議題包括法律文件分類、法律文件摘要、類似過去案例搜尋、協助判刑等。本論文探討「強盜罪」與「恐嚇取財罪」的分類以及此兩種罪的刑期預測，我們針對「強盜罪」與「恐嚇取財罪」定義的法律要素，並嘗試自動擷取法律要素。研究中證明利用定義的法律要素確實改善案件分類以及量刑預測，最後討論「強盜罪」與「恐嚇取財罪」的特徵以及影響判刑的因素。

This paper aims at classifying robbery and intimidation and predicting their sentencing by considering defined legal factors. We introduce a framework to fetch legal factors of robbery and intimidation and use factors as features for case classification and sentencing prediction. It proves that legal factors indeed improve the results of case classification and sentencing prediction. We then discuss the influence of these legal factors in case classification and in sentencing prediction.