Named Entity Recognition in Twitter using Images and Text

TL;DR

This paper introduces a novel multi-level architecture for Twitter NER that leverages both images and text features, avoiding reliance on handcrafted features or domain-specific rules, and achieves competitive results.

Contribution

The paper presents a new multi-level model that combines image and text features for Twitter NER, reducing dependency on linguistic resources.

Findings

Achieved 0.59 F-measure on Ritter dataset

Demonstrated competitive performance without handcrafted features

Showed potential for improved NER in noisy social media texts

Abstract

Named Entity Recognition (NER) is an important subtask of information extraction that seeks to locate and recognise named entities. Despite recent achievements, we still face limitations with correctly detecting and classifying entities, prominently in short and noisy text, such as Twitter. An important negative aspect in most of NER approaches is the high dependency on hand-crafted features and domain-specific knowledge, necessary to achieve state-of-the-art results. Thus, devising models to deal with such linguistically complex contexts is still challenging. In this paper, we propose a novel multi-level architecture that does not rely on any specific linguistic resource or encoded rule. Unlike traditional approaches, we use features extracted from images and text to classify named entities. Experimental tests against state-of-the-art NER for Twitter on the Ritter dataset present competitive results (0.59 F-measure), indicating that this approach may lead towards better NER models.