On the Impact of Entity Linking in Microblog Real-Time Filtering

TL;DR

This paper investigates how Entity Linking enhances real-time microblog filtering by enriching post representations, demonstrating significant improvements in retrieval effectiveness through Wikipedia-based EL methods.

Contribution

It introduces the integration of Wikipedia-based Entity Linking into a state-of-the-art microblog filtering system, showing notable performance gains.

Findings

EL significantly improves filtering effectiveness

Wikipedia-based EL outperforms non-EL methods

Enhanced retrieval accuracy in real-time filtering

Abstract

Microblogging is a model of content sharing in which the temporal locality of posts with respect to important events, either of foreseeable or unforeseeable nature, makes applica- tions of real-time filtering of great practical interest. We propose the use of Entity Linking (EL) in order to improve the retrieval effectiveness, by enriching the representation of microblog posts and filtering queries. EL is the process of recognizing in an unstructured text the mention of relevant entities described in a knowledge base. EL of short pieces of text is a difficult task, but it is also a scenario in which the information EL adds to the text can have a substantial impact on the retrieval process. We implement a start-of-the-art filtering method, based on the best systems from the TREC Microblog track realtime adhoc retrieval and filtering tasks , and extend it with a Wikipedia-based EL method. Results show that the use of EL significantly improves over non-EL based versions of the filtering methods.