TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Open-Domain Question Answering	ELI5	E-MCA	Rouge-L	24.0	# 4
Open-Domain Question Answering	ELI5	E-MCA	Rouge-1	30.0	# 2
Open-Domain Question Answering	ELI5	E-MCA	Rouge-2	5.8	# 3

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/using-local-knowledge-graph-construction-to/open-domain-question-answering-on-eli5)](https://paperswithcode.com/sota/open-domain-question-answering-on-eli5?p=using-local-knowledge-graph-construction-to)`

Using Local Knowledge Graph Construction to Scale Seq2Seq Models to Multi-Document Inputs

IJCNLP 2019 · Angela Fan, Claire Gardent, Chloe Braud, Antoine Bordes ·

Query-based open-domain NLP tasks require information synthesis from long and diverse web results. Current approaches extractively select portions of web text as input to Sequence-to-Sequence models using methods such as TF-IDF ranking. We propose constructing a local graph structured knowledge base for each query, which compresses the web search information and reduces redundancy. We show that by linearizing the graph into a structured input sequence, models can encode the graph representations within a standard Sequence-to-Sequence setting. For two generative tasks with very long text input, long-form question answering and multi-document summarization, feeding graph representations as input can achieve better performance than using retrieved text portions.