Enhancing Cross-border EU E-commerce through Machine Translation: Needed Language Resources, Challenges and Opportunities

14647

Authors

nan

Abstract

AbstractThis paper discusses the role that statistical machine translation (SMT) can play in the development of cross-border EU e-commerce,by highlighting extant obstacles and identifying relevant technologies to overcome them. In this sense, it firstly proposes a typology of e-commerce static and dynamic textual genres and it identifies those that may be more successfully targeted by SMT. The specific challenges concerning the automatic translation of user-generated content are discussed in detail. Secondly, the paper highlights the risk of data sparsity inherent to e-commerce and it explores the state-of-the-art strategies to achieve domain adequacy via adaptation. Thirdly, it proposes a robust workflow for the development of SMT systems adapted to the e-commerce domain by relying on inexpensive methods. Given the scarcity of user-generated language corpora for most language pairs, the paper proposes to obtain monolingual target-language data to train language models and aligned parallel corpora to tune and evaluate MT systems by means of crowdsourcing.

Keywords

nan

Full Content

Enhancing Cross-border EU E-commerce through Machine Translation: Needed Language Resources, Challenges and Opportunities Meritxell Fernández-Barrera1, Vladimir Popescu1, Antonio Toral2, Federico Gaspari2, Khalid Choukri1 1ELDA/ELRA, 9 Rue des Cordelières, 75013 Paris, France 2ADAPT Centre, School of Computing, Dublin City University, Dublin 9, Ireland Email: {meritxell; vladimir; choukri}@elda.org, {atoral; fgaspari}@computing.dcu.ie Abstract This paper discusses the role that statistical machine translation (SMT) can play in the development of cross-border EU e-commerce, by highlighting extant obstacles and identifying relevant technologies to overcome them. In this sense, it firstly proposes a typology of e-commerce static and dynamic textual genres and it identifies those that may be more successfully targeted by SMT. The specific challenges concerning the automatic translation of user-generated content are discussed in detail. Secondly, the paper highlights the risk of data sparsity inherent to e-commerce and it explores the state-of-the-art strategies to achieve domain adequacy via adaptation. Thirdly, it proposes a robust workflow for the development of SMT systems adapted to the e-commerce domain by relying on inexpensive methods. Given the scarcity of user-generated language corpora for most language pairs, the paper proposes to obtain monolingual target-language data to train language models and aligned parallel corpora to tune and evaluate MT systems by means of crowdsourcing. Keywords: Statistical Machine Translation (SMT), e-commerce, user-generated content (UGC), domain adaptation, crowdsourcing 1. Lowering the Language Barrier to Strengthen Cross-border EU E-commerce One of the challenges currently hindering cross-border EU e-commerce is the language barrier. Some studies have indeed shown that online consumers prefer home market suppliers and suppliers with whom they share the same culture and language (Cardona & Martens, 2014; Gómez- Herrera et al., 2014). The association eCommerce Europe (2015: 31) reports for instance the difficulties faced by merchants in culturally and linguistically diverse countries to reach the whole potential market. Ensuring the multilingualism of e-commerce sites thus appears to be key for the gradual development of cross-border EU e- commerce. Nevertheless, given the costs associated with human translation and web localization services, entering the multilingual scene is not affordable for all types of enterprises. There is indeed still a percentage of enterprises for which setting up the infrastructure to undertake e-commerce is perceived as too costly compared to potential profits. In this scenario, a pan- European platform enhancing multilingual technologies for EU languages may reduce the costs of MT system development and thus become an economic asset for enterprises wanting to enter cross-border e-commerce. The benefits that could be reaped from this market enlargement may attract enterprises so far reluctant to such an investment. SMEs, which typically have limited resources available for language technology development, could be particularly impacted by the availability of such a platform. And given the weight of SMEs in EU’s economy (28% of EU28 GDP, Muller et al., 2014: 14), their increased participation in e-sales would have a non- negligible effect on EU e-commerce growth. 2. From Internationalization and Localization to MT: a Typology of E-commerce Textual Genres E-commerce sites are made up of a variety of textual types that …

Content truncated to 500 words

Labeled Text

AbstractThis paper discusses the role that statistical machine translation (SMT) can play in the development of cross-border EU e-commerce,by highlighting extant obstacles and identifying relevant technologies to overcome them. In this sense, it firstly proposes a typology of e-commerce static and dynamic textual genres and it identifies those that may be more successfully targeted by SMT. The specific challenges concerning the automatic translation of user-generated content are discussed in detail. Secondly, the paper highlights the risk of data sparsity inherent to e-commerce and it explores the state-of-the-art strategies to achieve domain adequacy via adaptation. Thirdly, it proposes a robust workflow for the development of SMT systems adapted to the e-commerce domain by relying on inexpensive methods. Given the scarcity of user-generated language corpora for most language pairs, the paper proposes to obtain monolingual target-language data to train language models and aligned parallel corpora to tune and evaluate MT systems by means of crowdsourcing.

Publication Details

Paper ID:
14647
Year:
2016
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan