Although general question answering has been well explored in recent years, temporal question answering is a task which has not received as much focus. Our work aims to leverage a popular approach used for general question answering, answer extractio
n, in order to find answers to temporal questions within a paragraph. To train our model, we propose a new dataset, inspired by SQuAD, a state-of-the-art question answering corpus, specifically tailored to provide rich temporal information by adapting the corpus WikiWars, which contains several documents on history's greatest conflicts. Our evaluation shows that a pattern matching deep learning model, often used in general question answering, can be adapted to temporal question answering, if we accept to ask questions whose answers must be directly present within a text.
Recently, domain shift, which affects accuracy due to differences in data between source and target domains, has become a serious issue when using machine learning methods to solve natural language processing tasks. With additional pretraining and fi
ne-tuning using a target domain corpus, pretraining models such as BERT (Bidirectional Encoder Representations from Transformers) can address this issue. However, the additional pretraining of the BERT model is difficult because it requires significant computing resources. The efficiently learning an encoder that classifies token replacements accurately (ELECTRA) pretraining model replaces the BERT pretraining method's masked language modeling with a method called replaced token detection, which improves the computational efficiency and allows the additional pretraining of the model to a practical extent. Herein, we propose a method for addressing the computational efficiency of pretraining models in domain shift by constructing an ELECTRA pretraining model on a Japanese dataset and additional pretraining this model in a downstream task using a corpus from the target domain. We constructed a pretraining model for ELECTRA in Japanese and conducted experiments on a document classification task using data from Japanese news articles. Results show that even a model smaller than the pretrained model performs equally well.
To build machine learning-based applications for sensitive domains like medical, legal, etc. where the digitized text contains private information, anonymization of text is required for preserving privacy. Sequence tagging, e.g. as done in Named Enti
ty Recognition (NER) can help to detect private information. However, to train sequence tagging models, a sufficient amount of labeled data are required but for privacy-sensitive domains, such labeled data also can not be shared directly. In this paper, we investigate the applicability of a privacy-preserving framework for sequence tagging tasks, specifically NER. Hence, we analyze a framework for the NER task, which incorporates two levels of privacy protection. Firstly, we deploy a federated learning (FL) framework where the labeled data are not shared with the centralized server as well as the peer clients. Secondly, we apply differential privacy (DP) while the models are being trained in each client instance. While both privacy measures are suitable for privacy-aware models, their combination results in unstable models. To our knowledge, this is the first study of its kind on privacy-aware sequence tagging models.
The research aims at conducting a detailec comparison between
teaching Arabic to Speakers of other languages for specific purposes
and daily life usage.The study also showed an explanation of the
most prominent concepts and common terminology in t
his field in
order to ensure the clarity of their implications for scholars in the
field of teaching Arabic to speakers of other languages.
Our aim of this paper is studying the problem on normal oscillations of system of capillary viscous fluids in vessel.
We prove results about the spectrum of the problem for rotating vessel and prove that the systems of root elements ( eigenelements
and associated elements ) form an Abel-Lidsky basis.
Also , we use some results from the theory of J-self adjoint operators in studying the spectrum of the problem for non-rotating vessel.
Empowerment is considered as one of the modern management practices,
which occupies a great space of administrators' and decision-makers'
thinking, due to its effects & positive impacts which extend from the
organization and its employees to th
e community as a whole.
Also ( good governance ) is one of the most important administrative
issue which has lots of attention at the current period, and its importance
stems from the results and the positive effects of it;
Where Governance ensures rational use of available resources, fight
against administrative corruption, achieve the objectives of the state
according to the best standards, which contribute to improve the standard
of living for individuals and contribute to eradicate poverty and achieve
social development.
The researcher has studied the relationship between the two concepts of
ex- empowerment and good governance through the study of the
relationship of each of them with the ethics and development. The
community of study was the banking sector, both public and private,
while the sample was an intentional sample. The researcher distributed
(313) questionnaires, and the most important results were:
1. Governance is Applied significantly in the private banking sector
more than in the public banking sector.
2. the principles of empowerment are applied in public banking sector
more than in private banking sector.
3. No statistically significant relationship between the empowerment of
workers and the degree of disclosure and transparency applied in
banks.
This study aims to explore the attitudes of university professors to the determining factors of employment satisfaction as regards salary, functional security, empowerment, relationships and social value.
To achieve the objectives of this study, the
researchers constructed a questionnaire consisting of 28 questions in order to collect primary data from the study sample, which consisted of 150 university professors at private and governmental universities.
The researcher distributed 150 questionnaires, of which 120 were returned.
These questionnaires were validated for statistical analysis.
The most important results were as follows:
- There are no significant differences between the members of the academic staff who work in the private or governmental universities in Syria concerning: salary and remunerations; relationships with colleagues; relationship with management, and empowerment.
- There are significant differences between them concerning: their
relationship with students; job security and social value.
- This study had also proved that the priorities of the academic staff about factors causative of satisfaction vary relating to the university sector (governmental or private). . .
A half diallel set of crosses among six inbred lines of sweet corn was evaluated to study heterosis and combining ability among plant height, ear height, ear diameter, number of rows per ear and ear yield per plant. The study was carried out at the a
gricultural research center in, GCSAR, Lattakia, Snoubar Jableh, during the 2010, 2011 seasons.
Result showed that almost all crosses expressed a significant positive heterosis effect for ear yield per plant relative to mid parents and better parents; whereas, the highest positive significant percentage of heterosis for ear yield per plant were expressed by the crosses (L4xL6) which gave (198.70%, 176.81%) and (L4xL6) which gave (196.94%, 168.56%), over mid parents and better parents, respectively.
The ratio (σ2GCA/σ2SCA) which was less than (1) showed that the non-additive gene action was more important than the additive gene action in all traits except plant height and ear height. The inbred lines L3 (17.061) and L4 (12.011) seemed to be the best general combiners for ear yield. Also, based on SCA effects, many of single crosses were identified as superior for ear yield, and the best hybrid was L3xL5(50.173).
