Multi-modal Retrieval of Tables and Texts Using Tri-encoder Models

Kostić, Bogdan; Risch, Julian; Möller, Timo

doi:10.18653/v1/2021.mrqa-1.8

Cited by 8 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ODTQA is also closely related to open-domain text-based QA (Kwiatkowski et al, 2019;Khashabi et al, 2021) and information retrieval (Luan et al, 2021a;Humeau et al, 2020;Tang et al, 2021). Compared to the text retrieval, the tabular characteristics need to be considered in ODTQA (Herzig et al, 2021;Chen et al, 2023;Kostic et al, 2021).…”

Section: Related Workmentioning

confidence: 99%

“…The uncased BERT-base model is used as the encoder on two datasets. Following Tri-encoder (Kostic et al, 2021) and UTP (Chen et al, 2023), no downprojection is used for the final representations on the NQ-TABLES dataset. However, for a comprehensive evaluation, we also reported the performance of the proposed model with projection in Table 7.…”

Section: B Experimental Setupmentioning

confidence: 99%

See 1 more Smart Citation

Enhancing Open-Domain Table Question Answering via Syntax- and Structure-aware Dense Retrieval

Jin,

Li,

Chen

et al. 2023

Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacifi

View full text Add to dashboard Cite

Open-domain table question answering aims to provide answers to a question by retrieving and extracting information from a large collection of tables. Existing studies of open-domain table QA either directly adopt text retrieval methods or consider the table structure only in the encoding layer for table retrieval, which may cause syntactical and structural information loss during table scoring. To address this issue, we propose a syntax-and structure-aware retrieval method for the open-domain table QA task. It provides syntactical representations for the question and uses the structural header and value representations for the tables to avoid the loss of fine-grained syntactical and structural information. Then, a syntactical-to-structural aggregator is used to obtain the matching score between the question and a candidate table by mimicking the human retrieval process. Experimental results show that our method achieves the state-of-the-art on the NQ-tables dataset and overwhelms strong baselines on a newly curated open-domain Text-to-SQL dataset 1 .

show abstract

Section: Related Workmentioning

confidence: 99%

Section: B Experimental Setupmentioning

confidence: 99%

Enhancing Open-Domain Table Question Answering via Syntax- and Structure-aware Dense Retrieval

Jin,

Li,

Chen

et al. 2023

Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacifi

View full text Add to dashboard Cite

show abstract

“…(2) Bi-Encoder (Kosti'c et al, 2021) is a dense retriever which uses a BERT-based encoder for questions, and a shared BERT-based encoder to separately en-code tables and text as representations for retrieval. (3) Tri-Encoder (Kosti'c et al, 2021) is a dense retriever that uses three individual BERT-based en-coder to separately encode questions, tables and text as representations.…”

Section: A1 Settingsmentioning

confidence: 99%

Reasoning over Hybrid Chain for Table-and-Text Open Domain QA

Zhong¹,

Huang²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Tabular and textual question answering requires systems to perform reasoning over heterogeneous information, considering table structure, and the connections among table and text. In this paper, we propose a ChAincentric Reasoning and Pre-training framework (CARP). CARP utilizes hybrid chain to model the explicit intermediate reasoning process across table and text for question answering.We also propose a novel chain-centric pretraining method, to enhance the pre-trained model in identifying the cross-modality reasoning process and alleviating the data sparsity problem. This method constructs the large-scale reasoning corpus by synthesizing pseudo heterogeneous reasoning paths from Wikipedia and generating corresponding questions. We evaluate our system on OTT-QA, a large-scale table-and-text open-domain question answering benchmark, and our system achieves the state-of-the-art performance. Further analyses illustrate that the explicit hybrid chain offers substantial performance improvement and interpretablity of the intermediate reasoning process, and the chain-centric pretraining boosts the performance on the chain extraction.

show abstract

“…Pan et al (2021) later follows this work and improves the table retrieval with a 2-step retriever. Kostić et al (2021) discusses the use of dense vector embeddings to enhance the performance of bi-and tri-encoder in retrieving both table and text.…”

Section: Introductionmentioning

confidence: 99%

End-to-End Table Question Answering via Retrieval-Augmented Generation

Pan¹,

Canim²,

Glaß³

et al. 2022

Preprint

View full text Add to dashboard Cite

Most existing end-to-end Table Question Answering (Table QA) models consist of a twostage framework with a retriever to select relevant table candidates from a corpus and a reader to locate the correct answers from table candidates. Even though the accuracy of the reader models is significantly improved with the recent transformer-based approaches, the overall performance of such frameworks still suffers from the poor accuracy of using traditional information retrieval techniques as retrievers. To alleviate this problem, we introduce T-RAG, an end-to-end Table QA model, where a non-parametric dense vector index is fine-tuned jointly with BART, a parametric sequence-to-sequence model to generate answer tokens. Given any natural language question, T-RAG utilizes a unified pipeline to automatically search through a table corpus to directly locate the correct answer from table cell. We apply T-RAG on recent open-domainTable QA benchmarks and demonstrate that the finetuned T-RAG model is able to achieve stateof-the-art performance in both the end-to-end Table QA and the table retrieval tasks.

show abstract

Multi-modal Retrieval of Tables and Texts Using Tri-encoder Models

Cited by 8 publications

References 20 publications

Enhancing Open-Domain Table Question Answering via Syntax- and Structure-aware Dense Retrieval

Enhancing Open-Domain Table Question Answering via Syntax- and Structure-aware Dense Retrieval

Reasoning over Hybrid Chain for Table-and-Text Open Domain QA

End-to-End Table Question Answering via Retrieval-Augmented Generation

Contact Info

Product

Resources

About