Improved Cross-Lingual Question Retrieval for Community Question Answering

Rücklé, Andreas; Swarnkar, Krishnkant; Gurevych, Iryna

doi:10.1145/3308558.3313502

Cited by 14 publications

(9 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They used those embeddings to translate query terms word by word into the document language. Rücklé et al (2019) trained NMT model for CLIR using out-domain data and synthetic data (created by translating in-domain monolingual English into German) to retrieve answers to German questions from English collection in the technical domain (AskUbuntu and StackOverflow).…”

Section: Related Workmentioning

confidence: 99%

Document Translation vs. Query Translation for Cross-Lingual Information Retrieval in the Medical Domain

Saleh¹,

Pecina²

2020

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

View full text Add to dashboard Cite

We present a thorough comparison of two principal approaches to Cross-Lingual Information Retrieval: document translation (DT) and query translation (QT). Our experiments are conducted using the cross-lingual test collection produced within the CLEF eHealth information retrieval tasks in 2013-2015 containing English documents and queries in several European languages. We exploit the Statistical Machine Translation (SMT) and Neural Machine Translation (NMT) paradigms and train several domain-specific and task-specific machine translation systems to translate the non-English queries into English (for the QT approach) and the English documents to all the query languages (for the DT approach). The results show that the quality of QT by SMT is sufficient enough to outperform the retrieval results of the DT approach for all the languages. NMT then further boosts translation quality and retrieval quality for both QT and DT for most languages, but still, QT provides generally better retrieval results than DT.

show abstract

Section: Related Workmentioning

confidence: 99%

Document Translation vs. Query Translation for Cross-Lingual Information Retrieval in the Medical Domain

Saleh¹,

Pecina²

2020

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

View full text Add to dashboard Cite

show abstract

“…This shows that our work can potentially benefit a much wider range of related tasks beyond duplicate question detection. For instance, future work could extend upon this by using our methods to obtain more training data in cross-lingual cQA setups (Joty et al, 2017;Rücklé et al, 2019b), or by combining them with other training strategies, e.g., using our methods for pre-training.…”

Section: Discussionmentioning

confidence: 99%

Neural Duplicate Question Detection without Labeled Training Data

Rücklé

Moosavi

Gurevych

2019

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conferen

Self Cite

View full text Add to dashboard Cite

Supervised training of neural models to duplicate question detection in community Question Answering (cQA) requires large amounts of labeled question pairs, which are costly to obtain. To minimize this cost, recent works thus often used alternative methods, e.g., adversarial domain adaptation. In this work, we propose two novel methods: (1) the automatic generation of duplicate questions, and (2) weak supervision using the title and body of a question. We show that both can achieve improved performances even though they do not require any labeled data. We provide comprehensive comparisons of popular training strategies, which provides important insights on how to 'best' train models in different scenarios. We show that our proposed approaches are more effective in many cases because they can utilize larger amounts of unlabeled data from cQA forums. Finally, we also show that our proposed approach for weak supervision with question title and body information is also an effective method to train cQA answer selection models without direct answer supervision.

show abstract

“…Most of them are collected from open-domain and comprise of comparable/parallel document pairs, e.g., WaCky translation dataset (Baroni et al 2009;Joulin et al 2018), benchmark CLEF corpora (Vulić and Moens 2015), the Askubuntu benchmark corpus in QA task (Dos Santos et al 2015;Barzilay et al 2016), Arabic-English language pairs (Da San Martino et al 2017), text stream alignment 2018) follow previous dataset construction method (Schamoni et al 2014) and collect 25 cross-lingual datasets with large scales based on Wikipedia. Rücklé et al (2019) extend open-domain crosslingual question retrieval to the task-oriented domain, i.e., constructing a dataset upon StackOverflow. To facilitate the development of cross-lingual information retrieval in crossborder e-Commerce, we, for the first time, create a highquality heuristic dataset from real commercial applications.…”

Section: Existing Datasetsmentioning

confidence: 99%

“…According to the translation direction, these systems are further categorized into translating the local language to the foreign one and translating the foreign language to the local one. By doing so, the CLIR task is converted to a monolingual setting through the machine translation system (Nie 2010;Rücklé, Swarnkar, and Gurevych 2019). Although these systems are conceptual simple and natural, they are restricted by the performance of machine translation system.…”

Section: Introductionmentioning

confidence: 99%

Cross-Lingual Low-Resource Set-to-Description Retrieval for Global E-Commerce

Liu

Wang

et al. 2020

AAAI

View full text Add to dashboard Cite

With the prosperous of cross-border e-commerce, there is an urgent demand for designing intelligent approaches for assisting e-commerce sellers to offer local products for consumers from all over the world. In this paper, we explore a new task of cross-lingual information retrieval, i.e., cross-lingual set-to-description retrieval in cross-border e-commerce, which involves matching product attribute sets in the source language with persuasive product descriptions in the target language. We manually collect a new and high-quality paired dataset, where each pair contains an unordered product attribute set in the source language and an informative product description in the target language. As the dataset construction process is both time-consuming and costly, the new dataset only comprises of 13.5k pairs, which is a low-resource setting and can be viewed as a challenging testbed for model development and evaluation in cross-border e-commerce. To tackle this cross-lingual set-to-description retrieval task, we propose a novel cross-lingual matching network (CLMN) with the enhancement of context-dependent cross-lingual mapping upon the pre-trained monolingual BERT representations. Experimental results indicate that our proposed CLMN yields impressive results on the challenging task and the context-dependent cross-lingual mapping on BERT yields noticeable improvement over the pre-trained multi-lingual BERT model.

show abstract

Improved Cross-Lingual Question Retrieval for Community Question Answering

Cited by 14 publications

References 33 publications

Document Translation vs. Query Translation for Cross-Lingual Information Retrieval in the Medical Domain

Document Translation vs. Query Translation for Cross-Lingual Information Retrieval in the Medical Domain

Neural Duplicate Question Detection without Labeled Training Data

Cross-Lingual Low-Resource Set-to-Description Retrieval for Global E-Commerce

Contact Info

Product

Resources

About