539 research outputs found
Sort by
Problémy dnešních generátorů jazyka
No full textNatural language generation is a task in which a computer uses some input data to produce a comprehensible and fluent message in, for example, English or Czech. How does it work
Constrained Decoding for Technical Term Retention in English-Hindi MT
No full textTechnical terms may require special handling
when the target audience is bilingual,
depending on the cultural and educational
norms of the society in question. In particular,
certain translation scenarios may require
“term retention” i.e.
preserving of the
source language technical terms in the target
language output to produce a fluent and
comprehensible codeswitched sentence.
We show that a standard transformerbased
machine translation model can be adapted
easily to perform this task with little or no
damage to the general quality of its output.
We present an EnglishtoHindi model that is
trained to obey a “retain” signal, i.e. it can
perform the required codeswitching on a list
of terms, possibly unseen, provided at runtime.
We perform automatic evaluation using BLEU
as well as F1 metrics on the list of retained
terms; we also collect manual judgments on
the quality of the output sentences
ELITR Minuting Corpus: A Novel Dataset for Automatic Minuting from Multi-Party Meetings in English and Czech
No full textTaking minutes is an essential component of every meeting, although the goals, style, and procedure of this activity (“minuting” for
short) can vary. Minuting is a relatively unstructured writing act and is affected by who takes the minutes and for whom the minutes
are intended. With the rise of online meetings, automatic minuting would be an important use-case for the meeting participants and
those who might have missed the meeting. However, automatically generating meeting minutes is a challenging problem due to
various factors, including the quality of automatic speech recognition (ASR), public availability of meeting data, subjective knowledge
of the minuter, etc. In this work, we present the first of its kind dataset on Automatic Minuting. We develop a dataset of English
and Czech technical project meetings, consisting of transcripts generated from ASRs, manually corrected, and minuted by several
annotators. Our dataset, ELITR Minuting Corpus, consists of 120 English and 59 Czech meetings, covering about 180 hours of
meeting content. The corpus is publicly available at http://hdl.handle.net/11234/1-4692 as a set of meeting transcripts
and minutes, excluding the recordings for privacy reasons. A unique feature of our dataset is that most meetings are equipped with
more than one minute, each created independently. Our corpus thus allows studying differences in what people find important while
taking the minutes. We also provide baseline experiments for the community to explore this novel problem further. To the best of our
knowledge, ELITR Minuting Corpus is probably the first resource on minuting in English and also in a language other than English (Czech)
CUNI Systems for the WMT 22 Czech-Ukrainian Translation Task
No full textWe present Charles University submissions to the WMT22 General Translation Shared Task on Czech-Ukrainian and Ukrainian-Czech machine translation. We present two constrained submissions based on block back-translation and tagged back-translation and experiment with rule-based romanization of Ukrainian. Our results show that romanization only has a minor effect on the translation quality. Further, we describe Charles Translator, a system developed in March 2022 as a response to the migration from Ukraine to the Czech Republic. Compared to our constrained systems, it did not use romanization and used some proprietary data sources
CLS Infra Computational Literary Studies Infrastructure
No full textComputational Literary Studies Infrastructure, funded by the Horizon2020 grant scheme, is a four-year, pan-European project that aims to unify the diverse landscape of computational text analysis, in terms of available texts, tools, methods, practices and so forth, within its growing international user community. The project started out in February 2021, meaning that it has been underway for just over a year. In our poster we discuss the various deliverables and activities that have come out of the CLS INFRA project in its first quarter to give an idea of its impact in practice
ALIGNMEET: A Comprehensive Tool for Meeting Annotation, Alignment, and Evaluation
No full textMeeting summarization is primarily focused on topi cal coverage rather than on fluency or coherence. It is a challenging and tedious task, even when meeting summaries are created manually. The resulting sum maries vary in the goals, style, and they are inevitably very subjective due to the human in the loop. Also, the awareness of the context of the meeting is essential to create adequate and informative summaries
Elitr LangTools and THEaiTRE script generation
No full textPresentation of Elitr language tools and THEaiTRE script generation for EUROSAI congress attendees
Multimodality for NLP-Centered Applications: Resources, Advances and Frontiers
No full textWith the development of multimodal systems and natural language generation techniques, the resurgence of multimodal datasets has attracted significant research interests, which aims to provide new information to enrich the representation of textual data. However, there remains a lack of a comprehensive survey for this task. To this end, we take the first step and present a thorough review of this research field. This paper provides an overview of a publicly available dataset with different modalities according to the applications. Furthermore, we discuss the new frontier and give our thoughts. We hope this survey of multimodal datasets can provide the community with quick access and a general picture of the multimodal dataset for specific Natural Language Processing (NLP) applications and motivates future researches. In this context, we release the collection of all multimodal datasets easily accessible here: https://github.com/drmuskangarg/Multimodal-dataset
Language Technologies for the Creation of Multilingual Terminologies. Lessons Learned from the SSHOC Project
No full textThis paper is framed in the context of the SSHOC project and aims at exploring how Language Technologies can help in promoting and facilitating multilingualism in the Social Sciences and Humanities (SSH). Although most SSH researchers produce culturally and societally relevant work in their local languages, metadata and vocabularies used in the SSH domain to describe and index research data are currently mostly in English. We thus investigate Natural Language Processing and Machine Translation approaches in view of providing resources and tools to foster multilingual access and discovery to SSH content across different languages. As case studies, we create and deliver as freely, openly available data a set of multilingual metadata concepts and an automatically extracted multilingual Data Stewardship terminology. The two case studies allow as well to evaluate performances of state-of-the-art tools and to derive a set of recommendations as to how best apply them. Although not adapted to the specific domain, the employed tools prove to be a valid asset to translation tasks. Nonetheless, validation of results by domain experts proficient in the language is an unavoidable phase of the whole workflow
Multilingual SynSemClass for the Semantic Web (MSSW)
No full textLLOD (Linguistic Linked Open Data) is a generic name for a set of mutually connected language resources, using ontological relations. The connections between concepts and between concepts and their expressions in natural language make them suitable for both research and industrial applications in the area of content analysis, natural language understanding, (language- and knowledge-based) inferencing and other tasks. In the presented task, the concrete work will be on converting the SynSemClass project dataset (in part as a result of a previous Humane AI Net microproject called META-O-NLU) into LLOD, connecting it to the huge amount or interlinked data already available. A partner is involved in the Prêt-à-LLOD H2020 project, making this project synergistic in nature and multiplicative in terms of results in previous projects. Partners are also involved in the COST Action “European network for Web-centered linguistic data science” (NexusLinguarum)