Constructing Semantic Knowledge Base based on Wikipedia automation

DOI

10.2991/meita-16.2017.43How to use a DOI?

Keywords

The learning based on positive example, Feature Engineering, Semantic Relation, entity extraction, Hierarchical Reasoning, Wikipedia

Abstract

We know that Wikipedia is the largest knowledge set in the world, each instance entries can be a semantic entity, and it has richly hyperlinked text. Based on these, we propose a self-training method based on a small number of positive examples to extract the semantic relations and entities from the dynamic construction of semantic knowledge base. At the same time, we use TFIDF in the field of text classification and Feature Engineering in the field of computer linguistics to extract the physical characteristics of each instance and calculate their correlation. These physical features are used to help improve the accuracy and recall rate of the self-training method based on a small number of positive examples. After getting the entity right, it will be stored in the form of XML. Based on the storage structure of the XML document, a new reasoning algorithm is proposed which we called Hierarchical Reasoning. We use Wikipedia XML data in 2007 as the data test set the experimental results show that the filter based on feature selection constraint can obtain high precision and recall rate. In general, the knowledge base is built automatically. This makes it possible to extracting a large amount information from Wikipedia.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Wanpeng Niu
AU  - Junting Chen
AU  - Meilin Chen
PY  - 2017/02
DA  - 2017/02
TI  - Constructing Semantic Knowledge Base based on Wikipedia automation
BT  - Proceedings of the 2016 2nd International Conference on Materials Engineering and Information Technology Applications (MEITA 2016)
PB  - Atlantis Press
SP  - 202
EP  - 209
SN  - 2352-5401
UR  - https://doi.org/10.2991/meita-16.2017.43
DO  - 10.2991/meita-16.2017.43
ID  - Niu2017/02
ER  -