The paper describes the 3 NMT models submitted by the eTranslation team to the WMT 2021 news translation shared task. We developed systems in language pairs that are actively used in the European Commission's eTranslation service. In the WMT news tas
k, recent years have seen a steady increase in the need for computational resources to train deep and complex architectures to produce competitive systems. We took a different approach and explored alternative strategies focusing on data selection and filtering to improve the performance of baseline systems. In the domain constrained task for the French--German language pair our approach resulted in the best system by a significant margin in BLEU. For the other two systems (English--German and English-Czech) we tried to build competitive models using standard best practices.
This paper describes Netmarble's submission to WMT21 Automatic Post-Editing (APE) Shared Task for the English-German language pair. First, we propose a Curriculum Training Strategy in training stages. Facebook Fair's WMT19 news translation model was
chosen to engage the large and powerful pre-trained neural networks. Then, we post-train the translation model with different levels of data at each training stages. As the training stages go on, we make the system learn to solve multiple tasks by adding extra information at different training stages gradually. We also show a way to utilize the additional data in large volume for APE tasks. For further improvement, we apply Multi-Task Learning Strategy with the Dynamic Weight Average during the fine-tuning stage. To fine-tune the APE corpus with limited data, we add some related subtasks to learn a unified representation. Finally, for better performance, we leverage external translations as augmented machine translation (MT) during the post-training and fine-tuning. As experimental results show, our APE system significantly improves the translations of provided MT results by -2.848 and +3.74 on the development dataset in terms of TER and BLEU, respectively. It also demonstrates its effectiveness on the test dataset with higher quality than the development dataset.
In this paper we present the FJWU's system submitted to the biomedical shared task at WMT21. We prepared state-of-the-art multilingual neural machine translation systems for three languages (i.e. German, Spanish and French) with English as target lan
guage. Our NMT systems based on Transformer architecture, were trained on combination of in-domain and out-domain parallel corpora developed using Information Retrieval (IR) and domain adaptation techniques.
In this paper, we report the experimental results of Machine Translation models conducted by a NECTEC team for the translation tasks of WAT-2021. Basically, our models are based on neural methods for both directions of English-Myanmar and Myanmar-Eng
lish language pairs. Most of the existing Neural Machine Translation (NMT) models mainly focus on the conversion of sequential data and do not directly use syntactic information. However, we conduct multi-source neural machine translation (NMT) models using the multilingual corpora such as string data corpus, tree data corpus, or POS-tagged data corpus. The multi-source translation is an approach to exploit multiple inputs (e.g. in two different formats) to increase translation accuracy. The RNN-based encoder-decoder model with attention mechanism and transformer architectures have been carried out for our experiment. The experimental results showed that the proposed models of RNN-based architecture outperform the baseline model for English-to-Myanmar translation task, and the multi-source and shared-multi-source transformer models yield better translation results than the baseline.
The research aims to study the reality of Marketing Intelligence (Customer Intelligence,
Competitor Intelligence, Communicating Marketing Information among departments) in
the private banks operating in the Syrian Coast. The research adapts the des
criptive
approach. A questionnaire has been developed to collect primary data from the research
sample, which is a soft sample of 50 of Syrian Coast Private Banks' managers. 39 valid
questionnaires have been returned, which then have been analyzed by using some
Statistical tools provided by the statistical program SPSS, like: Means, relative importance,
and T-test. The most important research's results are represented in:The Banks under study
adapt customer intelligence as one of the most important components of marketing
intelligence, with a relative importance reached to (75.81%), and Competitor Intelligence
as one of the most important components of marketing intelligence, with a relative
importance reached to (73.846%).The researcher also made a number of recommendations,
including: The banks in question should give a more formal character to the activities of
marketing intelligence, carry out pilot studies to obtain information about customers, and
They should have update and detailed reports about competitors; to stay informed about
their strategies and movements.
This study aimed to investigate the impact of Human Resource Management Practices on the Employees performance in jude ҆s company ,these 7 Practices are (Planning, analysis and job design, polarization and recruitment, training, performance appraisal, motivation, participation in decision-making) and its impact on employees performance.
Housing is considered void that protects the rights of the external factors and provide
him with the necessary requirements, and depending on the quality and the pioneers of
global research in this range, the most important measure of the quality o
f the residential
building is the degree of satisfaction achieved by the building of the user , Since users are
the first to be affected by the decision of the house's design to understand they have the
right to participate in decision-making in the design process.
The research presents a field study on the posts in the design of residential buildings
in the city of Latakia in order to check the impact of user participation on the quality of the
residential building designed by measuring the relationship between the degree of user
participation and the amount of satisfaction from their homes.
The research found that the user has the participation of a major role in raising the
quality of the design of the apartment building in the city of Latakia in line with available
resources.
أتت فكرة المشروع من الأهمية المتزايدة للنظم المفتوحة المصدر في أيامنا هذه لاسيما الإمكانات الواسعة التي تتيحها هذه النظم في مجال إدارة الشبكات, حيث يهدف مشروعنا إلى إظهار مزايا نظام Ubuntu وذلك من خلال عرض وإعداد مجموعة من الخدمات التي يقدها في مجال
إدارة الشبكات, وبالتالي إظهار الفائدة العلمية والعملية منها, حيث نرى الجانب العلمي من خلال شرح ماتقوم به كل خدمة وماهي البروتوكولات والآليات التي تبنى عليها الخدمة, والتي أيضاً تظهر بشكل واضح من خلال الجانب العملي لكل خدمة لمافيه من عرض شامل للفائدة التي يمكن الحصول عليها.
This study aims to compare between two globally approved strategies to manage diabetes type II in Syria, by comparing their effectiveness in patients with diabetes type II who are treated in the national program of diabetes Clinics in Damascus and in
vestigate their glucose control. The first strategy includes glibenclamide as a monotherapy and the second includes a combination of glibenclamide and metformin.
The advantage of peer-to-peer (P2P) paradigm relies on two main concepts: cooperation among
users and resource sharing. There are many applications based on peer-to-peer paradigm, but
the most popular one is the file sharing. We can classify the fi
le sharing application into
centralized systems, (having a central server), and decentralized systems. Another classification
would be structured and unstructured systems, based on the way of managing the indexing
information. In this paper, we have implemented a centralized peer-to-peer application for file
sharing. Then we evaluated the performance of the system by means of simulation.
