Okapi+QuEst: Translation Quality Estimation within Okapi

نویسندگان

  • Gustavo Henrique Paetzold
  • Lucia Specia
  • Yves Savourel
چکیده

Due to the ever growing applicability of machine translation, estimating the quality of translations automatically has become a necessary task in various scenarios, for example, when deciding whether a machine translation is good enough for human post-editing. This demonstration presents the outcome of a collaborative project between the University of Sheffield and ENLASO, funded by EAMT, the European Association for Machine Translation. The project aimed to integrate a lightweight and user-friendly version of QuEst (http://www.quest.dcs.shef.ac.uk/) – a quality estimation toolkit, into Okapi (http://www.opentag.com/okapi/) – a framework with various components and applications designed to help create and improve translation and localisation processes. As result, Okapi users are now offered a software plugin to build and apply quality estimation models for translations produced within the framework. In addition to the standard functionalities of QuEst, the project involved the creation of new methods to facilitate the generation of the linguistic resources necessary for building quality estimation models.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Passage Retrieval Using Answer Type Profiles in Question Answering

Retrieving answer containing passages is a challenging task in Question Answering. In this paper, we describe a novel passage retrieval methodology using answer type profiles. Our methodology includes two steps: estimation and ranking. In the estimation step, answer type profiles are constructed from question-answer sentence pairs parallel corpus using a statistical alignment model. Each answer...

متن کامل

Okapi: Causally Consistent Geo-Replication Made Faster, Cheaper and More Available

Okapi is a new causally consistent geo-replicated keyvalue store. Okapi leverages two key design choices to achieve high performance. First, it relies on hybrid logical/physical clocks to achieve low latency even in the presence of clock skew. Second, Okapi achieves higher resource efficiency and better availability, at the expense of a slight increase in update visibility latency. To this end,...

متن کامل

Distinct and Diverse: Range-Wide Phylogeography Reveals Ancient Lineages and High Genetic Variation in the Endangered Okapi (Okapia johnstoni)

The okapi is an endangered, evolutionarily distinctive even-toed ungulate classified within the giraffidae family that is endemic to the Democratic Republic of Congo. The okapi is currently under major anthropogenic threat, yet to date nothing is known about its genetic structure and evolutionary history, information important for conservation management given the species' current plight. The d...

متن کامل

University of Hagen at CLEF 2005: Towards a Better Baseline for NLP Methods in Domain-Specific Information Retrieval

The third participation of the University of Hagen at the German Indexing and Retrieval Test (GIRT) task of the Cross Language Evaluation Campaign (CLEF 2005) aims at providing a better baseline for experiments with natural language processing (NLP) methods in domainspecific information retrieval (IR). Our monolingual experiments with the German document collection are based on a setup combinin...

متن کامل

Evolutionary analysis of vision genes identifies potential drivers of visual differences between giraffe and okapi

BACKGROUND The capacity of visually oriented species to perceive and respond to visual signal is integral to their evolutionary success. Giraffes are closely related to okapi, but the two species have broad range of phenotypic differences including their visual capacities. Vision studies rank giraffe's visual acuity higher than all other artiodactyls despite sharing similar vision ecological de...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2015