Papers by Constantin Orasan
What do Large Language Models Need for Machine Translation Evaluation? (2024.emnlp-main)
Copied to clipboard
Shenbin Qian, Archchana Sindhujan, Minnie Kabra, Diptesh Kanojia, Constantin Orasan, Tharindu Ranasinghe, Fred Blain
| Challenge: | Existing research shows that large language models can perform better in machine translation tasks. |
| Approach: | They propose to use large language models for machine translation evaluations . authors explore what translation information is needed for LLMs to evaluate MT quality . |
| Outcome: | The proposed model performs comparable to fine-tuned multilingual pre-trained models. |
A Semi-Automated Live Interlingual Communication Workflow Featuring Intralingual Respeaking: Evaluation and Benchmarking (2022.lrec-1)
Copied to clipboard
| Challenge: | Traditionally, live interlingual communication has been achieved only with the help of human interpreters. |
| Approach: | They propose a semi-automated workflow which uses a human respeaker and speaker-dependent speech recognition software to deliver punctuated same-language output of superior quality than the out-of-the-box ASR system. |
| Outcome: | The proposed workflow produces a similar quality output to the best-in-class simultaneous interpreters working with the same source speeches from the European Parliament. |
Automatic Linking of Judgements to UK Supreme Court Hearings (2023.emnlp-industry)
Copied to clipboard
| Challenge: | a number of legal documents are archived in the UK, including the Supreme Court's decisions and video recordings of court hearings. |
| Approach: | They propose to link segments in the text judgement to semantically relevant timespans in the videos of the hearings. |
| Outcome: | The proposed tool links segments in the text judgement to semantically relevant timespans in the videos of the hearings. |
MUSTS: MUltilingual Semantic Textual Similarity Benchmark (2025.acl-short)
Copied to clipboard
| Challenge: | Existing benchmarks for semantic textual similarity (STS) are limited to high-resource languages and do not include datasets annotated focusing on relatedness instead of similarity. |
| Approach: | They propose to evaluate multilingual semantic textual similarity benchmarks which span 13 languages and annotated datasets to evaluate and compare them. |
| Outcome: | The proposed method is the most comprehensive benchmark of multilingual STS methods. |
An Exploratory Analysis of Multilingual Word-Level Quality Estimation with Cross-Lingual Transformers (2021.acl-short)
Copied to clipboard
| Challenge: | Existing word-level quality estimation models require labelled data for each language pair and expensive maintenance. |
| Approach: | They propose to use multilingual QE models to generalise across languages . they propose to train models on other language pairs to predict word-level quality . |
| Outcome: | The proposed models generalise well across languages, making them more useful in real-world scenarios. |
Character-level Language Models for Abbreviation and Long-form Detection (2024.lrec-main)
Copied to clipboard
| Challenge: | Abbreviations and long forms are textual elements that are present in scientific communication . non-recognition of abbreviation and long form can lead to a negative impact on information retrieval . |
| Approach: | They propose to train and test language models for automatically identifying abbreviations and long forms . they use existing datasets annotated with abbrevations and their associated long forms to test them . |
| Outcome: | The proposed model can detect abbreviations and long forms on biomedical data . the proposed model improves on a previously untested dataset with biomedically-annotated datasets . |
TransQuest: Translation Quality Estimation with Cross-lingual Transformers (2020.coling-main)
Copied to clipboard
| Challenge: | Recent advances in the field of sentence-level quality estimation (QE) are based on neural-based architectures that require resourceintensive training. |
| Approach: | They propose a framework for sentence-level quality estimation based on cross-lingual transformers and use it to implement and evaluate two different neural architectures. |
| Outcome: | The proposed framework outperforms open-source QE frameworks when trained on WMT datasets and is very competitive in transfer learning settings. |
Linking Judgement Text to Court Hearing Videos: UK Supreme Court as a Case Study (2024.lrec-main)
Copied to clipboard
| Challenge: | Typically, transcripts of legal hearings are lengthy, making it time-consuming for legal professionals to analyse crucial arguments. |
| Approach: | They propose to use judgement-hearing pairs to link sections of written judgements with relevant moments in Supreme Court hearing videos to improve access to justice. |
| Outcome: | The proposed tool connects sections of written judgements with relevant moments in Supreme Court hearing videos, streamlining access to critical information. |
Centrality-aware Product Retrieval and Ranking (2024.emnlp-industry)
Copied to clipboard
| Challenge: | Ambiguity and complexity of user queries often lead to mismatch between user’s intent and retrieved product titles or documents. |
| Approach: | They propose a user-intent centrality optimization approach which optimizes for the user intent in semantic product search. |
| Outcome: | The proposed approach improves product ranking efficiency for ambiguous queries and lexical terms with alphanumeric characters. |