WikiPassageQA

Cohen, Daniel; Liu, Yang; Croft, W. Bruce

doi:10.1145/3209978.3210118

Cited by 39 publications

(9 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is in line with recent work in ad-hoc retrieval, which showed that BERT models trained on tweets and Wikipedia data transfer surprisingly well to news articles (Akkalyoncu Yilmaz et al, 2019). Other work has shown that IR baselines are often hard to beat, e.g., most neural models trained in-domain on WikiPassageQA perform below BM25 (Cohen et al, 2018). In contrast, we show that a large number of BERT models from a variety of 140 domains outperform these baselines without requiring any in-domain supervision.…”

Section: Resultssupporting

confidence: 87%

“…Typically, one answer is correct. • WikiPassageQA (Cohen et al, 2018) was crowd-sourced from Wikipedia articles and is not restricted to a particular domain (although many questions are about history topics). Candidate answers are passages from a single doc-ument, on the basis of which the question was formulated.…”

Section: Answer Selection (As)mentioning

confidence: 99%

“…Zero-shot transfer of text matching models is particularly challenging in setups of non-factoid answer selection (Cohen et al, 2018;Tay et al, 2017;Feng et al, 2015;Verberne et al, 2010) and question similarity (Nakov et al, 2017;Lei et al, 2016). These tasks compare questions and answers, or two potentially related questions in community question answering (cQA) forums, FAQ pages, and general collections of text passages.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

MultiCQA: Zero-Shot Transfer of Self-Supervised Text Matching Models on a Massive Scale

Rücklé

Pfeiffer

Gurevych

2020

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

View full text Add to dashboard Cite

We study the zero-shot transfer capabilities of text matching models on a massive scale, by self-supervised training on 140 source domains from community question answering forums in English. We investigate the model performances on nine benchmarks of answer selection and question similarity tasks, and show that all 140 models transfer surprisingly well, where the large majority of models substantially outperforms common IR baselines. We also demonstrate that considering a broad selection of source domains is crucial for obtaining the best zero-shot transfer performances, which contrasts the standard procedure that merely relies on the largest and most similar domains. In addition, we extensively study how to best combine multiple source domains. We propose to incorporate self-supervised with supervised multi-task learning on all available source domains. Our best zero-shot transfer model considerably outperforms in-domain BERT and the previous state of the art on six benchmarks. Fine-tuning of our model with in-domain data results in additional large gains and achieves the new state of the art on all nine benchmarks.

show abstract

Section: Resultssupporting

confidence: 87%

Section: Answer Selection (As)mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

MultiCQA: Zero-Shot Transfer of Self-Supervised Text Matching Models on a Massive Scale

Rücklé

Pfeiffer

Gurevych

2020

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

View full text Add to dashboard Cite

show abstract

“…Two datasets were used to evaluate this model. First with WikiPassageQA (Cohen et al, 2018) consists of 4165 question-answer pairs. This dataset was created from the Wikipedia text and the questions were created by crowd sourcing.…”

Section: Findings Datasetmentioning

confidence: 99%

A Novel Framework Using Zero Shot Learning Technique for a Non-Factoid Question Answering System

Devi

Sathick

Khan

et al. 2021

International Journal of Web-Based Learning and Teaching Technologies

View full text Add to dashboard Cite

Non-Factoid Question Answering (QA) is the next generation of textual QA systems, which gives passage level summaries for a natural language query, posted by the user. The main issue lies in the appropriateness of the generated summary. This paper proposes a framework for non-factoid QA system, which has three main components: (i) A deep neural network classifier, which produces sentence vector considering word correlation and context. (ii) Zero shot classifier that uses a multi-channel Convolutional Neural Network (CNN), to extract knowledge from multiple sources in the knowledge accumulator. This output acts as a knowledge enhancer that strengthens the passage level summary. (iii) Summary generator that uses Maximal Marginal Relevance (MMR) algorithm, which computes similarity among the query related answer and the sentences from zero shot classifier. This model is applied on the datasets WikiPassageQA and ANTIQUE. The experimental analysis shows that this model gives comparatively better results for WikiPassageQA dataset.

show abstract

“…Question answering Although they may not be not strictly formulated as questions, retrieval for complex topics is related to approaches that answer questions from web content. Retrieval techniques for effective question answering is undergoing a resurgence of research, with a particular interest in non-factoid QA [2], [3]. These works are similar to complex answer retrieval in that they perform question answering by retrieving relevant passages, in particular paragraphs from Wikipedia.…”

Section: Background and Related Workmentioning

confidence: 99%

Local and Global Query Expansion for Hierarchical Complex Topics

Dalton

Naseri

Dietz

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In this work we study local and global methods for query expansion for multifaceted complex topics. We study word-based and entity-based expansion methods and extend these approaches to complex topics using fine-grained expansion on different elements of the hierarchical query structure. For a source of hierarchical complex topics we use the TREC Complex Answer Retrieval (CAR) benchmark data collection. We find that leveraging the hierarchical topic structure is needed for both local and global expansion methods to be effective. Further, the results show that entity-based expansion methods show significant gains over word-based models alone, with local feedback providing the largest improvement. The results on the CAR paragraph retrieval task demonstrate that expansion models that incorporate both the hierarchical query structure and entity-based expansion result in a greater than 20% improvement over word-based expansion approaches.

show abstract

WikiPassageQA

Cited by 39 publications

References 15 publications

MultiCQA: Zero-Shot Transfer of Self-Supervised Text Matching Models on a Massive Scale

MultiCQA: Zero-Shot Transfer of Self-Supervised Text Matching Models on a Massive Scale

A Novel Framework Using Zero Shot Learning Technique for a Non-Factoid Question Answering System

Local and Global Query Expansion for Hierarchical Complex Topics

Contact Info

Product

Resources

About