A Verb-based Algorithm for Multiple-Relation Extraction from Single Sentences

Hao, Q., Keppens, J. and Rodrigues, O.

Proceedings of the 16th International Conference on Information and Knowledge Engineering. 115-121.

July 2017

Abstract

With the growing amount of unstructured articles written in natural-language, automated extracting knowledge of associations between entities is becoming essential for many applications. In this paper, we developed an automated verb-based algorithm for multiple-relation extraction from unstructured data obtained on-line. Named Entity Recognition (NER) techniques were applied to extract biomedical entities and relations were recognized by algorithms with Natural Language Processing (NLP) techniques. Evaluation based on F-measure with random sample of sentences from biomedical literature resulted an average precision of 90% and recall of 82%. We also compared the performance of the proposed algorithm against single-relation extraction algorithm, indicating improvements of this work. In conclusion, the preliminary study indicates that this method for multiple-relation extraction from unstructured literature is effective. With different training dataset, the algorithm can be applied to different domains. In addition, the automated method can be applied to detect and predict hidden relationships among varying areas.

Download PDF document