arXivDaily arXiv每日学术速递 周一至周五更新
arXiv cs/0408061cs.CL
收藏

电子词典作为NLP工具的基础:希腊案例

An electronic dictionary as a basis for NLP tools: The Greek case

Ch. Tsalidis, A. Vagelatos, G. Orphanos

摘要:

如果要在形态和句法层面上处理现代希腊语,则必须有一本电子表格形式的现代希腊语词典,因为现代希腊语是一种高度屈折的语言,具有显著的重音和拼写系统,具有许多古希腊语的特征。此外,如果要创建具有大量语言支持和覆盖范围的高效和复杂的NLP应用程序,则需要这样的工具。本论文将集中在这样一个现代希腊语电子词典的部署,这是内置的两个阶段:首先,它被构造为拼写校正模式的基础,然后它被重建,以成为一个更广泛的自然语言处理工具的部署平台。

英文摘要:

The existence of a Dictionary in electronic form for Modern Greek (MG) is mandatory if one is to process MG at the morphological and syntactic levels since MG is a highly inflectional language with marked stress and a spelling system with many characteristics carried over from Ancient Greek. Moreover, such a tool becomes necessary if one is to create efficient and sophisticated NLP applications with substantial linguistic backing and coverage. The present paper will focus on the deployment of such an electronic dictionary for Modern Greek, which was built in two phases: first it was constructed to be the basis for a spelling correction schema and then it was reconstructed in order to become the platform for the deployment of a wider spectrum of NLP tools.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑