PORE: Positive-Only Relation Extraction from Wikipedia Text

Abstract

Abstract. Extracting semantic relations is of great importance for the creation of the Semantic Web content. It is of great benefit to semi-automatically extract relations from the free text of Wikipedia using the structured content readily available in it. Pattern matching methods that employ information redundancy cannot work well since there is not much redundancy information in Wikipedia, compared to the Web. Multi-class classification methods are not reasonable since no classification of relation types is available in Wikipedia. In this paper, we propose PORE (Positive-Only Relation Extraction), for relation extraction from Wikipedia text. The core algorithm B-POL extends a state-of-the-art positive-only learning algorithm using bootstrapping, strong negative identification, and transductive inference to work with fewer positive training examples. We conducted experiments on several relations with different amount of training data. The experimental results show that B-POL can work effectively given only a small amount of positive training examples and it significantly outperforms the original positive learning approaches and a multi-class SVM. Furthermore, although PORE is applied in the context of Wikipedia, the core algorithm B-POL is a general approach for Ontology Population and can be adapted to other domains.

Semantic Relationship Extraction and Ontology Building using Wikipedia: A Comprehensive Survey

“…Therefore, it is highly preferable to use infoboxes them combined with one or more methods. To the contrary, KYLIN [7][8] used the information available on Wikipedia to complete and create infoboxes that belongs to the same class and sometimes it benefitted from the attributes in the infoboxes of some articles to add more attributes to other articles' infoboxes within the same class. As mentioned in [19] 44.2% of Wikipedia articles have infoboxes while categories covered nearly 81% of Wikipedia information.…”

“…Similarly, Wu and Weld [7] proposed the KYLIN system, which automatically extracts information from Wikipedia articles that belongs to the same category in order to create and complete their infoboxes. KYLIN generates infoboxes using Infoboxes Generation Module which consists of three phases: 1) preprocessing, 2) classifying and 3) extracting.…”

“…Based on KYLIN [7], Wu and Weld [8] proposed KOG (KYLIN Ontology Generator), which generally used to build Ontology that combines Wikipedia infoboxes and WordNet hierarchy. In KOG each infobox template was considered as a class, while the slots belong to the infobox were considered as attributes.…”

Disambiguating identity web references using Web 2.0 data and semantics

“…10: end while 11: return P and N According to [46] and [49] ranking instances based on their classification confidences and then enlarging the relevant training sets with the top k instances yields tradeoffs between resulting accuracy and optimized performance. Work in [49] states that lowering k achieves high accuracy whereas increasing k speeds up the process.…”

“…Work in [49] states that lowering k achieves high accuracy whereas increasing k speeds up the process. Therefore we must set k to a suitable value given the domain of application.…”