Interactive distributed corpus exploration and annotation infrastructure for large corpora and knowledge-bases

适用于大型语料库和知识库的交互式分布式语料库探索和注释基础设施

• 批准号: 315979217
• 负责人: Dr.-Ing. Richard Eckart de Castilho
• 金额: --
• 依托单位: Fachgebiet Ubiquitäre Wissensverarbeitung
• 依托单位国家: 德国
• 项目类别: Research data and software (Scientific Library Services and Information Systems)
• 财政年份: 2016
• 资助国家: 德国
• 起止时间: 2015-12-31 至 2021-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/315979217?language=en
• 关键词: Interactive; distributed; corpus; exploration; annotation

The goal of this project is a research infrastructure for corpus annotation that scales to large text document collections by flexibly building subcorpora. The infrastructure addresses the needs of computational linguists and corpus linguists for a generic tool to perform selective semantic annotation tasks within and across documents. Such an infrastructure is important because it enables the targeted exploitation of the huge amounts of digitally available text for linguistic analysis. The expert user should be supported by the infrastructure in exploring the large document collections, in setting up an annotation scheme, and in extracting task-specific subcorpora from a large background corpus. The annotation of the corpora should be flexibly distributable to remotely working annotation teams of different qualification levels and backgrounds. Their work should be supported through prioritisation and annotation suggestions based on machine learning technology to efficiently create a large corpus with high-quality annotations for training and evaluating the respective algorithms. Thus, infrastructure should enable the annotation of the same corpus from multiple perspectives by multiple researchers and annotations teams working in parallel. Custom corpora should be importable by the users as needed. Further functionality is needed to maintain and expand the knowledge bases used during the semantic annotation tasks as well as to connect to external standard knowledge bases.

这个项目的目标是一个语料库注释的研究基础设施，通过灵活地构建子语料库来扩展到大型文本文档集合。该基础设施解决了计算语言学家和语料库语言学家对通用工具的需求，以在文档内和跨文档执行选择性语义注释任务。这种基础设施非常重要，因为它能够有针对性地利用大量数字文本进行语言分析。专家用户应该支持的基础设施，在探索大型文档集，在建立一个注释计划，并在提取特定任务的子语料库从一个大的背景语料库。 语料库的注释应该可以灵活地分发给不同资格级别和背景的远程工作注释团队。他们的工作应该通过基于机器学习技术的优先级和注释建议来支持，以有效地创建具有高质量注释的大型语料库，用于训练和评估相应的算法。因此，基础设施应该能够从多个角度对同一语料库进行注释，由多个研究人员和注释团队并行工作。自定义语料库应该是可导入的用户根据需要。需要进一步的功能来维护和扩展语义注释任务期间使用的知识库，以及连接到外部标准知识库。