A Data-Intensive Approach to Named Entity Recognition Combining Contextual and Intrinsic Indicators

摘要

Over the past decade, huge volumes of valuable information have become available to organizations. However, the existence of a substantial part of the information in unstructured form makes the automated extraction of business intelligence and decision support information from it difficult. By identifying the entities and their roles within unstructured text in a process known as semantic named entity recognition, unstructured text can be made more readily available for traditional business processes. The authors present a novel NER approach that is independent of the text language and subject domain making it applicable within different organizations. It departs from the natural language and machine learning methods in that it leverages the wide availability of huge amounts of data as well as high-performance computing to provide a data-intensive solution. Also, it does not rely on external resources such as dictionaries and gazettes for the language or domain knowledge.