Cross-document Event Coreference Search: Task, Dataset and Modeling (2022-10-23T00:00:00.000000Z)

TL;DR

A novel model is presented that integrates a powerful coreference scoring scheme into the DPR architecture, yielding improved performance and adapting the prominent Deep Passage Retrieval model to the setting, as an appealing baseline.

Abstract

The task of Cross-document Coreference Resolution has been traditionally formulated as requiring to identify all coreference links across a given set of documents. We propose an appealing, and often more applicable, complementary set up for the task – Cross-document Coreference Search, focusing in this paper on event coreference. Concretely, given a mention in context of an event of interest, considered as a query, the task is to find all coreferring mentions for the query event in a large document collection. To support research on this task, we create a corresponding dataset, which is derived from Wikipedia while leveraging annotations in the available Wikipedia Event Coreferecene dataset (WEC-Eng). Observing that the coreference search setup is largely analogous to the setting of Open Domain Question Answering, we adapt the prominent Deep Passage Retrieval (DPR) model to our setting, as an appealing baseline. Finally, we present a novel model that integrates a powerful coreference scoring scheme into the DPR architecture, yielding improved performance.

Authors

Ido Dagan

19 papers

Alon Eirew

6 papers

Avi Caciularu

6 papers

TL;DR

Abstract

Authors

References52 items

Contrastive Representation Learning for Cross-Document Coreference Resolution of Events and Entities

The Web Is Your Oyster - Knowledge-Intensive NLP against a Very Large Web Corpus

Unsupervised Dense Information Retrieval with Contrastive Learning

Salient Phrase Aware Dense Retrieval: Can a Dense Retriever Imitate a Sparse One?

Focus on what matters: Applying Discourse Coherence Theory to Cross Document Coreference

Sequential Cross-Document Coreference Resolution

BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models

Efficiently Teaching an Effective Dense Retriever with Balanced Topic Aware Sampling

WEC: Deriving a Large-scale Cross-document Event Coreference dataset from Wikipedia

Decontextualization: Making Sentences Stand-Alone

Generalizing Cross-Document Event Coreference Resolution Across Multiple Corpora

Paired Representation Learning for Event and Entity Coreference

Streamlining Cross-Document Coreference Resolution: Evaluation and Modeling

Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval

Paraphrasing vs Coreferring: Two Sides of the Same Coin

ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT

Dense Passage Retrieval for Open-Domain Question Answering

CorefQA: Coreference Resolution as Query-based Span Prediction

Do Multi-hop Readers Dream of Reasoning Chains?

BERT for Coreference Resolution: Baselines and Analysis

Natural Questions: A Benchmark for Question Answering Research

RoBERTa: A Robustly Optimized BERT Pretraining Approach

SpanBERT: Improving Pre-training by Representing and Predicting Spans

Coreference Resolution with Entity Equalization

Latent Retrieval for Weakly Supervised Open Domain Question Answering

Revisiting Joint Modeling of Cross-document Entity and Event Coreference Resolution

End-to-End Open-Domain Question Answering with BERTserini

Abstract Meaning Representation for Multi-Document Summarization

Don’t Annotate, but Validate: a Data-to-Text Method for Capturing Event Data

Higher-Order Coreference Resolution with Coarse-to-Fine Inference

Neural Models for Reasoning over Multiple Mentions Using Coreference

Decoupled Weight Decay Regularization

Concept-Map-Based Multi-Document Summarization using Concept Coreference Resolution and Global Importance Optimization

End-to-end Neural Coreference Resolution

TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension

Reading Wikipedia to Answer Open-Domain Questions

SQuAD: 100,000+ Questions for Machine Comprehension of Text

Modeling of the Question Answering Task in the YodaQA System

Using a sledgehammer to crack a nut? Lexical diversity and event coreference resolution

Semantic Parsing on Freebase from Question-Answer Pairs

A Multi-Pass Sieve for Coreference Resolution

The Probabilistic Relevance Framework: BM25 and Beyond

Event Coreference Data (Almost) for Free: Mining Hyperlinks from Online News

CDLM: Cross-document language modeling

KBPearl: A Knowledge Base Population System Supported by Joint Entity and Relation Linking

2020). Additionally, our training procedure depends on challenging negative examples provided by a BM25 model (§4.1)

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

48th Annual Meeting of the Association for Computational Linguistics

was an officer in the Wehrmacht of Nazi Germany during World War II who was responsible for German military operations causing Cephalonia Massacre

The TREC-8 Question Answering Track Report

He resumed his former business pursuits, and was a delegate to the 1860 Republican National Convention that nominated Abraham Lincoln as the candidate for president

2017. spaCy 2: Natural language understanding with Bloom embeddings, convolutional neural networks and incremental parsing

Field of Study

Venue Information

Name

Type

URL

Alternate Names