Search

A dual-band wearable antenna is designed on a textile material. The design operates at ISM bands available for Wireless Body Area Network and Wireless Local Area Network with an input match better than -15 dB. The antenna is designed by using Computational Electromagnetic Software (CEMS) based on finite-difference time-domain (FDTD) method. A three-layer phantom model including skin, fat and muscle has been considered to compute the specific absorption rate (SAR). The maximum value of SAR averaged over 1g and 10g of tissue is less than 1.6 W/Kg and 2 W/Kg, respectively, when the maximum incident power of the antenna is 63 mW. These values are incompliance with the international electromagnetic safety standards.

Transfer learning from pre-trained neural language models towards downstream tasks has been a predominant theme in NLP recently. Several researchers have shown that deep NLP models learn non-trivial amount of linguistic knowledge, captured at different layers of the model. We investigate how fine-tuning towards downstream NLP tasks impacts the learned linguistic knowledge. We carry out a study across popular pre-trained models BERT, RoBERTa and XLNet using layer and neuron-level diagnostic classifiers. We found that for some GLUE tasks, the network relies on the core linguistic information and preserve it deeper in the network, while for others it forgets. Linguistic information is distributed in the pre-trained language models but becomes localized to the lower layers post fine-tuning, reserving higher layers for the task specific knowledge. The pattern varies across architectures, with BERT retaining linguistic information relatively deeper in the network compared to RoBERTa and XLNet, where it is predominantly delegated to the lower layers., Comment: Findings of the ACL 2021

This paper is a write-up for the tutorial on "Fine-grained Interpretation and Causation Analysis in Deep NLP Models" that we are presenting at NAACL 2021. We present and discuss the research work on interpreting fine-grained components of a model from two perspectives, i) fine-grained interpretation, ii) causation analysis. The former introduces methods to analyze individual neurons and a group of neurons with respect to a language property or a task. The latter studies the role of neurons and input features in explaining decisions made by the model. We also discuss application of neuron analysis such as network manipulation and domain adaptation. Moreover, we present two toolkits namely NeuroX and Captum, that support functionalities discussed in this tutorial., Comment: Accepted at NAACL Tutorial

While a lot of analysis has been carried to demonstrate linguistic knowledge captured by the representations learned within deep NLP models, very little attention has been paid towards individual neurons.We carry outa neuron-level analysis using core linguistic tasks of predicting morphology, syntax and semantics, on pre-trained language models, with questions like: i) do individual neurons in pre-trained models capture linguistic information? ii) which parts of the network learn more about certain linguistic phenomena? iii) how distributed or focused is the information? and iv) how do various architectures differ in learning these properties? We found small subsets of neurons to predict linguistic tasks, with lower level tasks (such as morphology) localized in fewer neurons, compared to higher level task of predicting syntax. Our study also reveals interesting cross architectural comparisons. For example, we found neurons in XLNet to be more localized and disjoint when predicting properties compared to BERT and others, where they are more distributed and coupled., Accepted in EMNLP 2020

Roman Urdu is an informal form of the Urdu language written in Roman script, which is widely used in South Asia for online textual content. It lacks standard spelling and hence poses several normalization challenges during automatic language processing. In this article, we present a feature-based clustering framework for the lexical normalization of Roman Urdu corpora, which includes a phonetic algorithm UrduPhone, a string matching component, a feature-based similarity function, and a clustering algorithm Lex-Var. UrduPhone encodes Roman Urdu strings to their pronunciation-based representations. The string matching component handles character-level variations that occur when writing Urdu using Roman script. The similarity function incorporates various phonetic-based, string-based, and contextual features of words. The Lex-Var algorithm is a variant of the k-medoids clustering algorithm that groups lexical variations of words. It contains a similarity threshold to balance the number of clusters and their maximum similarity. The framework allows feature learning and optimization in addition to the use of predefined features and weights. We evaluate our framework extensively on four real-world datasets and show an F-measure gain of up to 15% from baseline methods. We also demonstrate the superiority of UrduPhone and Lex-Var in comparison to respective alternate algorithms in our clustering framework for the lexical normalization of Roman Urdu.

Pre-trained Transformer-based models have achieved state-of-the-art performance for various Natural Language Processing (NLP) tasks. However, these models often have billions of parameters, and, thus, are too resource-hungry and computation-intensive to suit low-capability devices or applications with strict latency requirements. One potential remedy for this is model compression, which has attracted a lot of research attention. Here, we summarize the research in compressing Transformers, focusing on the especially popular BERT model. In particular, we survey the state of the art in compression for BERT, we clarify the current best practices for compressing large-scale Transformer models, and we provide insights into the workings of various methods. Our categorization and analysis also shed light on promising future research directions for achieving lightweight, accurate, and generic NLP models., To appear in TACL 2021. The arXiv version is a pre-MIT Press publication version

Transformer-based deep NLP models are trained using hundreds of millions of parameters, limiting their applicability in computationally constrained environments. In this paper, we study the cause of these limitations by defining a notion of Redundancy, which we categorize into two classes: General Redundancy and Task-specific Redundancy. We dissect two popular pretrained models, BERT and XLNet, studying how much redundancy they exhibit at a representation-level and at a more fine-grained neuron-level. Our analysis reveals interesting insights, such as: i) 85% of the neurons across the network are redundant and ii) at least 92% of them can be removed when optimizing towards a downstream task. Based on our analysis, we present an efficient feature-based transfer learning procedure, which maintains 97% performance while using at-most 10% of the original neurons., 19 Pages, 14 figures, EMNLP 2020

The widespread usage of Twitter during emergencies has provided a new opportunity and timely resource to crisis responders for various disaster management tasks. Geolocation information of pertinent tweets is crucial for gaining situational awareness and delivering aid. However, the majority of tweets do not come with geoinformation. In this work, we focus on the task of location mention recognition from crisis-related tweets. Specifically, we investigate the influence of different types of labeled training data on the performance of a BERT-based classification model. We explore several training settings such as combing in- and out-domain data from news articles and general-purpose and crisis-related tweets. Furthermore, we investigate the effect of geospatial proximity while training on near or far-away events from the target event. Using five different datasets, our extensive experiments provide answers to several critical research questions that are useful for the research community to foster research in this important direction. For example, results show that, for training a location mention recognition model, Twitter-based data is preferred over general-purpose data; and crisis-related data is preferred over general-purpose Twitter data. Furthermore, training on data from geographically-nearby disaster events to the target event boosts the performance compared to training on distant events.

Low-resource machine translation suffers from the scarcity of training data and the unavailability of standard evaluation sets. While a number of research efforts target the former, the unavailability of evaluation benchmarks remain a major hindrance in tracking the progress in low-resource machine translation. In this paper, we introduce AraBench, an evaluation suite for dialectal Arabic to English machine translation. Compared to Modern Standard Arabic, Arabic dialects are challenging due to their spoken nature, non-standard orthography, and a large variation in dialectness. To this end, we pool together already available Dialectal Arabic-English resources and additionally build novel test sets. AraBench offers 4 coarse, 15 fine-grained and 25 city-level dialect categories, belonging to diverse genres, such as media, chat, religion and travel with varying level of dialectness. We report strong baselines using several training settings: fine-tuning, back-translation and data augmentation. The evaluation suite opens a wide range of research frontiers to push efforts in low-resource machine translation, particularly Arabic dialect translation. The evaluation suite and the dialectal system are publicly available for research purposes.

Two sets of novel extensions of NNJM model are proposed. The NDAM models that regularizes the loss function with respect to in-domain model, give an improvement of up to +0.4 BLEU points. The NFM models that fuse in- and out-domain NNJM models give an improvement of up to +0.9 BLEU points. The NFM models also beat state-of-the-art phrase-table adaptation methods. The gains obtained from NNJM and phrase-table adaptation were found to be additive. We explore neural joint models for the task of domain adaptation in machine translation in two ways: (i)we apply state-of-the-art domain adaptation techniques, such as mixture modelling and data selection using the recently proposed Neural Network Joint Model (NNJM) (Devlin etal., 2014); (ii)we propose two novel approaches to perform adaptation through instance weighting and weight readjustment in the NNJM framework. In our first approach, we propose a pair of models called Neural Domain Adaptation Models (NDAM) that minimizes the cross entropy by regularizing the loss function with respect to in-domain (and optionally to out-domain) model. In the second approach, we present a set of Neural Fusion Models (NFM) that combines the in- and the out-domain models by readjusting their parameters based on the in-domain data.We evaluated our models on the standard task of translating English-to-German and Arabic-to-English TED talks. The NDAM models achieved better perplexities and modest BLEU improvements compared to the baseline NNJM, trained either on in-domain or on a concatenation of in- and out-domain data. On the other hand, the NFM models obtained significant improvements of up to +0.9 and +0.7 BLEU points, respectively. We also demonstrate improvements over existing adaptation methods such as instance weighting, phrasetable fill-up, linear and log-linear interpolations.

We present the theoretical development of a simple Receiver Autonomous Integrity Monitoring (RAIM) algorithm by modifying a Conventional Snapshot RAIM algorithm and its implementation on DSP platform. Integrity describes a navigation system capability to be responsible for well-timed warnings to users as soon as the satellite or whole constellation should not be used for navigation. RAIM algorithms detect, identify those satellite anomalies and failures and calculates the Protection Level (PL). PL is the 1east measurable position error along with the certain possibilities of missed detection and a false alert. The developed algorithm uses conventional snapshot based RAIM algorithm with some changes in the threshold measurements for the protection level calculations, which is then compared with a tested predefined threshold to provide precise positioning and a high level of integrity due to setting a very tight fault detection threshold and Protection Level. Final1y, the developed algorithm is implemented on DSP TMS320C641XX keeping the computational/processing requirements for RAIM Algorithm as low as possible and simulations are performed to compare the results.

Tools to measure lung and airways volume are critical for pulmonary researchers interested in evaluating the impact of disease or novel therapies on the lung. Barometric plethysmography is a classic technique to evaluate the lung volume with a long history of clinical use. Volumetric capnography utilizes the profile of exhaled carbon dioxide to determine the volume of the conducting airways, or dead space, and provides an index of airways homogeneity. These techniques may be used independently, or in combination to evaluate the dependence of airways volume and homogeneity on lung volume. This paper provides detailed technical instructions to replicate these techniques and our representative data demonstrates that the airways volume and homogeneity are highly correlated to lung volume. We also provide a macro for the analysis of capnographic data, which can be modified or adapted to fit different experimental designs. The advantage of these measures is that their advantages and limitations are supported by decades of experimental data, and they can be made repeatedly in the same subject without expensive imaging equipment or technically advanced analysis algorithms. These methods may be particularly useful for investigators interested in perturbations that change both the functional residual capacity of the lung and airways volume.

We share the findings of the first shared task on improving robustness of Machine Translation (MT). The task provides a testbed representing challenges facing MT models deployed in the real world, and facilitates new approaches to improve models; robustness to noisy input and domain mismatch. We focus on two language pairs (English-French and English-Japanese), and the submitted systems are evaluated on a blind test set consisting of noisy comments on Reddit and professionally sourced translations. As a new task, we received 23 submissions by 11 participating teams from universities, companies, national labs, etc. All submitted systems achieved large improvements over baselines, with the best improvement having +22.33 BLEU. We evaluated submissions by both human judgment and automatic evaluation (BLEU), which shows high correlations (Pearson's r = 0.94 and 0.95). Furthermore, we conducted a qualitative analysis of the submitted systems using compare-mt, which revealed their salient differences in handling challenges in this task. Such analysis provides additional insights when there is occasional disagreement between human judgment and BLEU, e.g. systems better at producing colloquial expressions received higher score from human judgment.

The significant part of smart grid is to make smart grid cost-efficient by predicting electricity price and load. To improve the prediction performance we proposed an Efficient Convolutional Neural Network (ECNN) and Efficient K-nearest Neighbour (EKNN) in which the parameters are tuned. It may be difficult to deal with huge amount of load data that is coming from the electricity market. To overcome this issue, we incorporated three modules in the proposed methodology. The proposed model consists of feature engineering and classification. Feature engineering is a two-step process (feature selection and feature extraction); for the purpose of feature selection Mutual Information (MI) is used which reduces the redundancy among features and for feature extraction Recursive Feature Elimination (RFE) is used to extract the principle features from the selected features and reduces the dimensionality of features. Finally, after training the data-set and the removal of the duplicate features load prediction is done by ECNN and EKNN. The ECNN and EKNN outperforms better then traditional Convolutional Neural Network (CNN) and K-nearest Neighbour (KNN). The forecast performance is evaluated by comparing the results with MAPE, RMSE, MAE and MSE. i.e. 10.8, 7.5, 7.15, and 10.4 respectively.

Arabic text is typically written without short vowels (or diacritics). However, their presence is required for properly verbalizing Arabic and is hence essential for applications such as text to speech. There are two types of diacritics, namely core-word diacritics and case-endings. Most previous works on automatic Arabic diacritic recovery rely on a large number of manually engineered features, particularly for case-endings. In this work, we present a unified character level sequence-to-sequence deep learning model that recovers both types of diacritics without the use of explicit feature engineering. Specifically, we employ a standard neural machine translation setup on overlapping windows of words (broken down into characters), and then we use voting to select the most likely diacritized form of a word. The proposed model outperforms all previous state-of-the-art systems. Our best settings achieve a word error rate (WER) of 4.49% compared to the state-of-the-art of 12.25% on a standard dataset.

Recent work has shown that contextualized word representations derived from neural machine translation are a viable alternative to such from simple word predictions tasks. This is because the internal understanding that needs to be built in order to be able to translate from one language to another is much more comprehensive. Unfortunately, computational and memory limitations as of present prevent NMT models from using large word vocabularies, and thus alternatives such as subword units (BPE and morphological segmentations) and characters have been used. Here we study the impact of using different kinds of units on the quality of the resulting representations when used to model morphology, syntax, and semantics. We found that while representations derived from subwords are slightly better for modeling syntax, character-based representations are superior for modeling morphology and are also more robust to noisy input.

We address the problem of simultaneous translation by modifying the Neural MT decoder to operate with dynamically built encoder and attention. We propose a tunable agent which decides the best segmentation strategy for a user-defined BLEU loss and Average Proportion (AP) constraint. Our agent outperforms previously proposed Wait-if-diff and Wait-if-worse agents (Cho and Esipova, 2016) on BLEU with a lower latency. Secondly we proposed data-driven changes to Neural MT training to better match the incremental decoding framework., NAACL 2018

A 1:16 Wilkinson power divider (WPD) working from 1.3 GHz to 3 GHz is presented in this paper. Equal power division, phase and amplitude consistency is achieved among the output ports. The proposed design has a wideband with a 10 dB fractional bandwidth of 76%. The isolation loss within the bandwidth is better than 15 dB and the insertion loss is less than 2 dB. The size of the WPD has been minimized to 16×10,7 cm 2, which was the main objective of this work. The compact size of the proposed design makes it a suitable candidate for feeding phased antenna arrays for wideband operations. IEEE IEEE Karachi Section Pakistan Science Foundation Institution of Engineers, Pakistan (IEP) Hamdard Foundation

Despite the remarkable evolution of deep neural networks in natural language processing (NLP), their interpretability remains a challenge. Previous work largely focused on what these models learn at the representation level. We break this analysis down further and study individual dimensions (neurons) in the vector representation learned by end-to-end neural models in NLP tasks. We propose two methods: Linguistic Correlation Analysis, based on a supervised method to extract the most relevant neurons with respect to an extrinsic task, and Cross-model Correlation Analysis, an unsupervised method to extract salient neurons w.r.t. the model itself. We evaluate the effectiveness of our techniques by ablating the identified neurons and reevaluating the network's performance for two tasks: neural machine translation (NMT) and neural language modeling (NLM). We further present a comprehensive analysis of neurons with the aim to address the following questions: i) how localized or distributed are different linguistic properties in the models? ii) are certain neurons exclusive to some properties and not others? iii) is the information more or less distributed in NMT vs. NLM? and iv) how important are the neurons identified through the linguistic correlation method to the overall task? Our code is publicly available as part of the NeuroX toolkit (Dalvi et al. 2019)., Comment: AAA 2019, pages 10, AAAI Conference on Artificial Intelligence (AAAI 2019)

In this work, we present Qatar Computing Research Institute»s live speech translation system. Our system works with both Arabic and English. It is designed using an array of modern web technologies to capture speech in real time, and transcribe and translate it using state-of-the-art Automatic Speech Recognition (ASR) and Machine Translation (MT) systems. The platform is designed to be useful in a wide variety of situations like lectures, talks and meetings. It is often the case in the Middle East that audiences in talks understand either Arabic or English alone. This system enables the speaker to talk in either language, and the audience to understand what is being spoken even if they are not bilingual.The system consists of three primary modules, i) a Web application, ii) ASR system, iii) and a statistical/neural MT system. The three modules are optimized to work jointly and process the speech at a real-time factor close to one - which means that the systems are optimized to keep up with the speaker and provide the results with a short delay, comparable to what we observe in (human) interpretation. The real-time factor for the entire pipeline is 1.18. The Web application is based on the standard HTML5 WebAudio application programming interface. It captures speech input from a microphone on the user»s device and transmits it to the backend servers for processing. The servers send back the transcriptions and translations of the speech, which is then displayed to the user. Our platform features a way to instantly broadcast live sessions for anyone to see the transcriptions and translations of a session in real-time without being physically present at the speaker»s location. The ASR system is based on KALDI, a state-of-the-art toolkit for speech recognition. We use a combination of time delay neural networks (TDNN) and long-short term memory neural network (LSTM) to ensure real time transcription of the incoming speech while ensuring high quality output. The Arabic and English systems have average word error rates of 23% and 9.7% respectively. The Arabic system consists of the following components: i) a character based lexicon of size 900K; the lexicon maps words to sound units to learn acoustic representation, ii) 40 dimensional high-resolution features extracted for each speech frame to digitize the audio signal, iii) a 100-dimensional i-vectors for each frame to facilitate speaker adaptation, iv) TDNN acoustic models, and v) Tri-gram language model trained using 110 M words, and restricted to 900 K vocabulary.The MT system has two choices for the backend – a statistical phrase-based system and a neural MT system. Our phrase-based system is trained with Moses, a state-of-the-art statistical MT framework, and the neural-based systems is trained with Nematus, a state-of-the-art neural MT framework. We use Modified Moore-Lewis filtering to select the best subset of the available data to train our phrase-based system more efficiently. In order to speed up the translation even further, we prune the language models backing the phrase-based system, ignoring knowledge that is not frequently used. On the other hand, our neural-based system MT system trained on all the available data as its training scales linearly with the amount of data unlike phrase-based systems. Our Neural MT system is roughly 3–5% better on the BLEU scale, a standard measure for computing the quality of translations. However, the existing neural MT decoders are slower than the phrase-based decoders translating 9.5 tokens/second versus 24 tokens/second. The trade-off between efficiency and accuracy barred us from picking only one final system. By enabling both technologies we allow the trade-off between quality and efficiency and leave it up to the user to decide whether they prefer fast or accurate system.Our system has been successfully demonstrated locally and globally at several venues like Al Jazeera, MIT, BBC and TII. The state-of-the-art technologies backing the platform for transcription and translation are also available independently and can be integrated seamlessly into any external platform. The Speech Translation system is publicly available at http://st.qcri.org/demos/livetranslation.

We present a toolkit to facilitate the interpretation and understanding of neural network models. The toolkit provides several methods to identify salient neurons with respect to the model itself or an external task. A user can visualize selected neurons, ablate them to measure their effect on the model accuracy, and manipulate them to control the behavior of the model at the test time. Such an analysis has a potential to serve as a springboard in various research directions, such as understanding the model, better architectural choices, model distillation and controlling data biases., Comment: AAAI Conference on Artificial Intelligence (AAAI 2019), Demonstration track, pages 2

Single Molecule Spectroscopy and Superresolution Imaging XI 2018 -- 27 January 2018 through 28 January 2018 -- 136394 This work presents a compact dual-band Planar Inverted F-antenna (PIFA) antenna useful for E-health monitoring and wireless sensors systems. The antenna operates in the Industrial Standard and Medical (ISM) and Wireless Medical Telemetry Service (WMTS) bands. It offers a compact size with dimensions 12.6 × 8.5 × 2.4 mm3. Two different simulators have been used to verify the results. The proposed antenna performs well in the presence of a bio-compatible insulator (BCI) material.

We present QCRI’s Arabic-to-English speech translation system. It features modern web technologies to capture live audio, and broadcasts Arabic transcriptions and English translations simultaneously. Our Kaldi-based ASR system uses the Time Delay Neural Network architecture, while our Machine Translation (MT) system uses both phrase-based and neural frameworks. Although our neural MT system is slower than the phrase-based system, it produces significantly better translations and is memory efficient.

Word segmentation plays a pivotal role in improving any Arabic NLP application. Therefore, a lot of research has been spent in improving its accuracy. Off-the-shelf tools, however, are: i) complicated to use and ii) domain/dialect dependent. We explore three language-independent alternatives to morphological segmentation using: i) data-driven sub-word units, ii) characters as a unit of learning, and iii) word embeddings learned using a character CNN (Convolution Neural Network). On the tasks of Machine Translation and POS tagging, we found these methods to achieve close to, and occasionally surpass state-of-the-art performance. In our analysis, we show that a neural machine translation system is sensitive to the ratio of source and target tokens, and a ratio close to 1 or greater, gives optimal performance., ACL 2017 pages 7

Neural machine translation (MT) models obtain state-of-the-art performance while maintaining a simple, end-to-end architecture. However, little is known about what these models learn about source and target languages during the training process. In this work, we analyze the representations learned by neural MT models at various levels of granularity and empirically evaluate the quality of the representations for learning morphology through extrinsic part-of-speech and morphological tagging tasks. We conduct a thorough investigation along several parameters: word-based vs. character-based representations, depth of the encoding layer, the identity of the target language, and encoder vs. decoder representations. Our data-driven, quantitative evaluation sheds light on important aspects in the neural MT system and its ability to capture word structure., Updated decoder experiments

Microblog classification has received a lot of attention in recent years. Different classification tasks have been investigated, most of them focusing on classifying microblogs into a small number of classes (five or less) using a training set of manually annotated tweets. Unfortunately, labelling data is tedious and expensive, and finding tweets that cover all the classes of interest is not always straightforward, especially when some of the classes do not frequently arise in practice. In this paper, we study an approach to tweet classification based on distant supervision, whereby we automatically transfer labels from one social medium to another for a single-label multi-class classification task. In particular, we apply YouTube video classes to tweets linking to these videos. This provides for free a virtually unlimited number of labelled instances that can be used as training data. The classification experiments we have run show that training a tweet classifier via these automatically labelled data achieves substantially better performance than training the same classifier with a limited amount of manually labelled data; this is advantageous, given that the automatically labelled data come at no cost. Further investigation of our approach shows its robustness when applied with different numbers of classes and across different languages.

We present an unsupervised method to find lexical variations in Roman Urdu informal text. Our method includes a phonetic algorithm UrduPhone, a featurebased similarity function, and a clustering algorithm Lex-C. UrduPhone encodes roman Urdu strings to their phonetic equivalent representations. This produces an initial grouping of different spelling variations of a word. The similarity function incorporates word features and their context. Lex-C is a variant of k-medoids clustering algorithm that group lexical variations. It incorporates a similarity threshold to balance the number of clusters and their maximum similarity. We test our system on two datasets of SMS and blogs and show an f-measure gain of up to 12% from baseline systems.

We present novel models for domain adaptation based on the neural network joint model (NNJM). Our models maximize the cross entropy by regularizing the loss function with respect to in-domain model. Domain adaptation is carried out by assigning higher weight to out-domain sequences that are similar to the in-domain data. In our alternative model we take a more restrictive approach by additionally penalizing sequences similar to the outdomain data. Our models achieve better perplexities than the baseline NNJM models and give improvements of up to 0.5 and 0.6 BLEU points in Arabic-to-English and English-to-German language pairs, on a standard task of translating TED talks.

In this paper, we take a closer look at the MT evaluation process from a glass-box perspective using eye-tracking. We analyze two aspects of the evaluation task ‐ the background of evaluators (monolingual or bilingual) and the sources of information available, and we evaluate them using time and consistency as criteria. Our findings show that monolinguals are slower but more consistent than bilinguals, especially when only target language information is available. When exposed to various sources of information, evaluators in general take more time and in the case of monolinguals, there is a drop in consistency. Our findings suggest that to have consistent and cost effective MT evaluations, it is better to use monolinguals with only target language information.

We describe the CMU-Q and QCRI’s joint efforts in building a spelling correction system for Arabic in the QALB 2015 Shared Task. Our system is based on a hybrid pipeline that combines rule-based linguistic techniques with statistical methods using language modeling and machine translation, as well as an error-tolerant finite-state automata method. We trained and tested our spelling corrector using the dataset provided by the shared task organizers. Our system outperforms the baseline system and yields better correction quality with an F-score of 68.12 on L1- test-2015 test set and 38.90 on the L2-test2015. This ranks us 2nd in the L2 subtask and 5th in the L1 subtask.

Several recent papers on Arabic dialect identification have hinted that using a word unigram model is sufficient and effective for the task. However, most previous work was done on a standard fairly homogeneous dataset of dialectal user comments. In this paper, we show that training on the standard dataset does not generalize, because a unigram model may be tuned to topics in the comments and does not capture the distinguishing features of dialects. We show that effective dialect identification requires that we account for the distinguishing lexical, morphological, and phonological phenomena of dialects. We show that accounting for such can improve dialect detection accuracy by nearly 10% absolute.

We demonstrate the feasibility of using unsupervised morphological segmentation for dialects of Arabic, which are poor in linguistics resources. Our experiments using a Qatari Arabic to English machine translation system show that unsupervised segmentation helps to improve the translation quality as compared to using no segmentation or to using ATB segmentation, which was especially designed for Modern Standard Arabic (MSA). We use MSA and other dialects to improve Qatari Arabic to English machine translation, and we show that a uniform segmentation scheme across them yields an improvement of 1.5 BLEU points over using no segmentation.

We investigate three methods for integrating an unsupervised transliteration model into an end-to-end SMT system. We induce a transliteration model from parallel data and use it to translate OOV words. Our approach is fully unsupervised and language independent. In the methods to integrate transliterations, we observedimprovements from 0.23-0.75 (∆ 0.41) BLEU points across 7 language pairs. Wealso show that our mined transliteration corpora provide better rule coverage andtranslation quality compared to the gold standard transliteration corpora.

In computer networks, file transfer methods are used as a tool to exchange files among computers. Efficiency of a file transfer method is generally measured in terms of bandwidth usage, integrity, security, scalability and fault tolerance. We have critically reviewed some state-of-the-art patent file transfer methods to know about their shortcomings and strengths for different scenarios. We analyzed that none of them is up to the mark in terms of all parameters of efficiency and there is a great need of research in this area. After a critical analysis, some methods for the improvement have also been proposed in this study.

In this paper, four state-of-art probabilistic taggers i.e. TnT tagger, TreeTagger, RF tagger and SVM tool, are applied to the Urdu language. For the purpose of the experiment, a syntactic tagset is proposed. A training corpus of 100,000 tokens is used to train the models. Using the lexicon extracted from the training corpus, SVM tool shows the best accuracy of 94.15%. After providing a separate lexicon of 70,568 types, SVM tool again shows the best accuracy of 95.66%.

We present a generative model that efficiently mines transliteration pairs in a consistent fashion in three different settings: unsupervised, semi-supervised, and supervised transliteration mining. The model interpolates two sub-models, one for the generation of transliteration pairs and one for the generation of non-transliteration pairs (i.e., noise). The model is trained on noisy unlabeled data using the EM algorithm. During training the transliteration sub-model learns to generate transliteration pairs and the fixed non-transliteration model generates the noise pairs. After training, the unlabeled data is disambiguated based on the posterior probabilities of the two sub-models. We evaluate our transliteration mining system on data from a transliteration mining shared task and on parallel corpora. For three out of four language pairs, our system outperforms all semi-supervised and supervised systems that participated in the NEWS 2010 shared task. On word pairs extracted from parallel corpora with fewer than 2% transliteration pairs, our system achieves up to 86.7% F-measure with 77.9% precision and 97.8% recall.

Short vowels, aka diacritics, are more often omitted when writing different varieties of Arabic including Modern Standard Arabic (MSA), Classical Arabic (CA), and Dialectal Arabic (DA). However, diacritics are required to properly pronounce words, which makes diacritic restoration (a.k.a. diacritization) essential for language learning and text-to-speech applications. In this paper, we present a system for diacritizing MSA, CA, and two varieties of DA, namely Moroccan and Tunisian. The system uses a character level sequence-to-sequence deep learning model that requires no feature engineering and beats all previous SOTA systems for all the Arabic varieties that we test on.

We present the first prototype of the SUMMA Platform: an integrated platform for multilingual media monitoring. The platform contains a rich suite of low-level and high-level natural language processing technologies: automatic speech recognition of broadcast media, machine translation, automated tagging and classification of named entities, semantic parsing to detect relationships between entities, and automatic construction / augmentation of factual knowledge bases. Implemented on the Docker platform, it can easily be deployed, customised, and scaled to large volumes of incoming media streams.