Supervised and Unsupervised Ensembling for Knowledge Base Population

TL;DR

This paper introduces a combined supervised and unsupervised ensembling approach that improves performance on Knowledge Base Population tasks, surpassing existing methods and demonstrating broad applicability.

Contribution

It proposes a novel ensembling method that integrates supervised and unsupervised techniques, outperforming previous systems in KBP tasks.

Findings

Outperforms the best systems in 2015 KBP competition

Surpasses several ensembling baselines and stacking methods

Demonstrates effectiveness on diverse KBP tasks

Abstract

We present results on combining supervised and unsupervised methods to ensemble multiple systems for two popular Knowledge Base Population (KBP) tasks, Cold Start Slot Filling (CSSF) and Tri-lingual Entity Discovery and Linking (TEDL). We demonstrate that our combined system along with auxiliary features outperforms the best performing system for both tasks in the 2015 competition, several ensembling baselines, as well as the state-of-the-art stacking approach to ensembling KBP systems. The success of our technique on two different and challenging problems demonstrates the power and generality of our combined approach to ensembling.