English-Turkish Shopping Domain Parallel Corpora

The dataset consists of bilingual sentence-aligned corpora for the Shopping domain from English to Turkish and vice versa.

About This OTS Dataset

Introduction

Welcome to the English-Turkish Bilingual Parallel Corpora dataset for the Shopping domain! This meticulously curated dataset offers a rich collection of bilingual text data, translated between English and Turkish, providing a valuable resource for developing Shopping domain-specific language models and machine translation engines.

Dataset Content

•Volume and Diversity:

•

Extensive Dataset: Over 50,000 sentences offering a robust dataset for various applications.

•

Translator Diversity: Contributions from more than 200 native translators ensure a wide range of linguistic styles and interpretations.

•Sentence Diversity:

•

Word Count: Sentences range from 7 to 25 words, suitable for various computational linguistic applications.

•

Syntactic Variety: The corpus encompasses sentences with varying syntactic structures, including simple, compound, and complex sentences.

•

Interrogative and Imperative Forms: The corpus includes sentences in interrogative (question) and imperative (command) forms, reflecting the conversational nature of the Shopping industry.

•

Affirmative and Negative Statements: Both affirmative and negative statements are represented in the corpus, ensuring different polarities.

•

Passive and Active Voice: The corpus features sentences written in both active and passive voice, ensuring different perspectives and representations of information.

•

Idiomatic Expressions and Figurative Language: The corpus incorporates idiomatic expressions, metaphors, and figurative language commonly used in the Shopping domain.

•

Discourse Markers and Connectives: The corpus includes a wide range of discourse markers and connectives, such as conjunctions, transitional phrases, and logical connectors, which are crucial for capturing the logical flow and coherence of the text.

•

Cross Translation: The dataset includes a cross-translation, where a part of the dataset is translated from English to Turkish and another portion is translated from Turkish to English, to improve bi-directional translation capabilities.

Domain Specific Content

This Parallel Corpus is meticulously curated to capture the linguistic intricacies and domain-specific nuances inherent to the Shopping industry.

•

Industry-Tailored Terminology: The corpus encompasses a comprehensive lexicon of Shopping-specific terminology, ranging from technical terms related to e-commerce, product descriptions, and payment processing to customer service and retail operations.

•

Authentic Industry Expressions: Beyond technical terminology, the corpus captures the authentic expressions, idioms, and colloquialisms used within the Shopping domain.

•

Contexts Specific to Shopping Domain: The corpus encompasses various shopping-related contexts, including product information such as descriptions and specifications, customer feedback in the form of reviews and ratings, transactional messages like payment confirmation and transaction updates, navigation and exploration aids like category descriptions and subcategory details, marketing materials including promotions, advertisements, and discounts, order management updates like tracking and shipping notifications, return and exchange policies and procedures, and customer support resources like FAQs and support sections.

•

Cross-Domain Applicability: While the corpus is specifically designed for the Shopping domain, it also includes relevant terminology and language from related areas, such as Fashion, Beauty, Electronics, Gadgets, etc

Format and Structure

•

Multiple Formats: Available in Excel format, with the ability to convert to JSON, TMX, XML, XLIFF, XLS, and other industry-standard formats, facilitating ease of use and integration.

•

Structure: It contains information like Serial Number, Unique ID, Source Sentence, Source Sentence Word Count, Target Sentence, and Target Sentence Word Count.

Usage and Application

•

Machine Translation: Develop accurate machine translation engines for shopping content localization, enabling seamless shopping experiences across languages.

•

NLP Applications: Enabling the creation and improvement of predictive keyboards, spell checkers, grammar checkers, and text/speech understanding systems.

•

Language Modeling: Train language models to generate product descriptions, reviews, and other shopping-related content.

Secure and Ethical Collection

•Our proprietary parallel corpus platform “Yugo” was used throughout the process of this dataset creation.

•Throughout the dataset creation process, the data remained within our secure platform and did not leave our environment, ensuring data security and confidentiality.

•It does not include any personally identifiable information, which makes the dataset safe to use.

•The source or translated content included in the corpus does not infringe upon any copyrights or intellectual property rights. The corpus comprises original content created specifically for this purpose.

Update and Customization

To ensure the continued relevance and effectiveness of this Shopping Domain Parallel Corpora Dataset for robust language models and machine translation engines, we are committed to regular updates.

•Customization & Custom Collection Options:

•

Annotation: Various types of annotations like Part-of-speech tagging, Named Entity Recognition (NER), Sentiment Analysis, Intent Classification, Multiple Translation Ranking, or any other application-specific annotations can be made available upon request.

•

Classification: Classification of corpus based on type of sentence, and subdomain can be made available.

•

Custom Collection: Custom collection can be done on specific requirements in any language pair and domain.