Abstract

Ever since secure multi-party computation was successfully showed [6], it has remained an interesting research field in the cryptography community. Multi-party computation is one of the most important primitive in modern cryptography, in which mutually distrusting parties want to compute a functionality over their private data without revealing any additional information about their data. It is well established that the information theoretic multi-party computation among n parties is possible if the number of corrupted nodes t is less than n 2 for semi-honest adversary and the number of corrupted nodes t is less than n 3 for malicious adversary. Despite the progress over past 3 decades, multi-party computation have many improvements to go through before it becomes viable for practical scenarios. Also, there happens to exist many areas which are little explored. In this dissertation we explore one such path where we rely on physical cryptographic assumption rather than the traditional assumptions. The traditional assumptions such as discrete log problem, RSA, etc. are always subject to non-existence of poly time algorithm for computational problems of certain nature. Also, these assumptions do not cover computationally unbounded adversary. The motivation to study such assumption is that the physical cryptographic assumptions can be characterized in a way that they are realistic as well as provide required security guarantees. In this thesis, we discover that channel asynchrony is one such assumption where the channel may not deliver the messages in the same order as they were sent. Usually we use solutions such as sequence number to order the messages in such case but for security such asynchrony happens to be a boon. It is also more efficient practically compared to the protocols in the information theoretic setting. This thesis also solves an open problem in the multi-party computation literature. “Is it possible to remove the dependency on the multiplicative depth of the circuit in the overall communication complexity of a secure MPC protocol while maintaining its efficiency and the restrictions on the number of corruptions such that 3ta + 2tp + tf < n?”. This work is based on the works of Goyal et al. and Hirt et al. Goyal et. al. solve a similar question for active adversarial setting instead of mixed adversary. Hirt et al. have also given the multiplicative depth dependent protocol for mixed adversarial setting. We combine these two works to answer the question stated above in affirmative. Middleboxes in a computer network system inspect and analyse network traffic to detect malicious communications, monitor system performance and provide operational services. However, encrypted traffic, which has become increasingly prevalent, hinders the ability of middleboxes to perform such services. A common practice in addressing this issue is by employing a “Man-in-the-Middle” (MitM)approach, wherein an encrypted traffic flow between two end-points is interrupted, decrypted and analyzed by the middleboxes. The MitM approach is straightforward and is used by many organisations, but there are both practical and privacy concerns. Practically, due to the cost of the MitM appliances and the latency incurred due to the encrypt-decrypt processes, enterprises continue to seek solutions that are less costly and less compute-intensive. There has also been discussion on the many efforts required to configure MitM. Besides, MitM violates end-to-end privacy guarantee, raising privacy concerns and potential issues on compliance especially with the rising awareness on user privacy. Furthermore, some of the MitM implementations were found to be flawed. Consequently, new practical and privacy-preserving techniques that enable inspection over encrypted traffic were proposed. In this thesis, we also systematically examine these techniques to compare their advantages, limitations and challenges. We categorise them into four main categories by defining a framework that consist of system architectures, use cases, trust and threat models. These are searchable encryption, access control, machine learning and trusted hardware. We first discuss the man-in-the-middle approach as a baseline, then discuss in details each of them, and provide an in-depth comparisons of their advantages and limitations. By doing so we describe practical constraints, advantages and pitfalls towards adopting the techniques. Following this, we give insights on the gaps between research work and practical implementation in the industries, which leads us to the discussion on the challenges and research directions.

Abstract

Natural Language Processing (NLP) systems have attracted substantial attention for their potential and propitious commercial value in recent years. Part-of-speech (POS) tagging is a well-studied task and is one of the essential and fundamental steps required for various NLP tasks. Numerous types of supervised, unsupervised, and semi-supervised POS taggers exist for different languages. Each of these POS taggers has its unique strengths. To leverage the combined benefits of these existing taggers, developing ensemble methods is a good option. Ensemble Methods are techniques that create multiple base models and then combine them to produce improved results and usually produce more accurate results than a single model. Our study investigates ensemble methods to exploit the complementary characteristics of Part-ofspeech taggers. We present an ensemble technique that enhances the performance of Supervised taggers with the help of Semi-supervised taggers. We propose a two-layer static ensemble with a decision tree classifier as the second layer. Outputs of the first level base POS taggers are chosen as new input features set to the second level decision tree. We use Semi-supervised context based lists (CBLs) tagger as one of the first layer taggers, which uses rich contextual information and helps in tagging both existing and unseen words and uses no domain knowledge, while supervised taggers give good performance for words present in the training model. Enhanced performance of our proposed ensemble approach over the base methods suggests that integrating these methods combines the strengths of base taggers. Later, the thesis focuses on developing a dynamic classifier selection approach suitable for POS tagging. In Dynamic Classifier Selection (DCS) techniques, the most appropriate set of classifiers is selected dynamically for each test sample. This selection is based on the performance of base classifiers on similar context words. In DCS methods, it is assumed that if a classifier performs well in the region of competence, that classifier is most likely to perform better for the given test instance. Word2Vec, a word Embedding algorithm trained over unlabeled data, is used to pick these similar context-based words from the validation set. Our proposed DCS approach on five languages outperforms the base POS taggers. These results suggest that we successfully selected the appropriate classifier for a given test word. Overall, we propose ensemble methods to address the Parts-of-speech tagging of resource-poor languages. These methods work with fewer annotated data by using unlabeled data without considering any domain-based knowledge. These ensemble methods can be an excellent choice to integrate and leverage the benefits of various classifiers and help the large number (6500+) of “resource-poor languages” that do not have much-annotated training data

Abstract

Machine Translation is the task where a machine generates a sentence in language T, given an input sentence in language S. The generated and input sentences must have similar semantics. In the current literature, there are various methods proposed to solve the task. Deep neural network based translation models (Neural Machine Translation (NMT)) have been shown to achieve state-of-the-art performance for multiple language pairs. However, neural methods struggle to perform well for lowresource languages. Low-resource languages are understudied, suffer from the data scarcity problem, and lack efficient language processing tools. Since neural methods require large amount of data for training effective models, less-complex statistical methods outperform these methods. Transformer-based NMT models have achieved state-of-the-art performance for various languages. Multiple parallel attention mechanisms that use multiple attention heads facilitate greater performance of the Transformer model for various applications e.g., NMT), text classification. However, transformer models do not perform well in low-resource conditions. In this thesis, we propose a novel Dynamic Head Importance Computation Mechanism (DHICM), which enhances the performance of the transformer model, especially, in low-resource conditions. In multi-head attention mechanism, different heads attend to different parts of the input. However, the limitation is that multiple heads might attend to the same part of the input, resulting in multiple heads being redundant. Thus, the model resources are underutilized. One approach to avoid this is to prune least important heads based on certain importance score. We focus on designing a head importance computation method to dynamically calculate the importance of a head with respect to the input. Our insight is to design an additional attention layer together with multi-head attention, and utilize the outputs of the multi-head attention along with the input, to compute the importance for each head. Additionally, we ad d an extra loss function to prevent the model from assigning same score to all heads, to identify more important heads and improvise performance. We analy zed performance of DHICM for NMT with different languages. Experiments on different datasets show that DHICM outperforms traditional Transformer-based approach by large margin, especially, when less training data is available. Code mixing is a phenomena of mixing two or more languages, prevalent in multilingual communities. In recent years, due to the prevalence of social media platforms, there has been a surge in the usage of code-mixed texts. Code-mixed texts are informal in nature, and do not necessarily follow pre-defined syntactic structures. Due to the informal nature, traditional NLP systems for monolingual languages, do not perform well for code-mixed inputs. Moreover, good quality code-mixed data is scarce, hence,

Abstract

Today human identification is a practice followed quite frequently in contexts ranging from forensic labs to high security organisations to almost every domestic household. One of the oldest and most accurate ways to establish the identity of an individual is through fingerprints. It is so because, biologically no two people (not even identical twins) have the same fingerprints. Hence it comes as no surprise that fingerprint biometrics has been a very active domain of research for a long time. Over the years, especially with the emergence of the deep learning era, the computational modules of a fingerprint recognition system have become bigger and bigger. Simultaneously, the actual hardware deploying fingerprint recognition systems has become more and more user friendly. These developments provide a contrast as one tries to incorporate the highly accurate but also extremely heavy modules of a fingerprint recognition system in the palm of our hands. In this thesis, we will perform a couple of studies which explore ways to optimize fingerprint recognition systems. This is done by trying to parallelize modules and eliminate semantically redundant computations. Along the way we will also obtain better performances as compared to the academic state of the art. Typical fingerprint recognition systems are comprised of a spoof detection module and a subsequent matching module, running one after the other. In the first study, we posit that both spoof detection and fingerprint matching are correlated tasks. Therefore, rather than performing the two tasks separately, we propose a joint model for spoof detection and matching to simultaneously perform both tasks without compromising the accuracy of either task. In practice, this reduces the time and memory requirements of the fingerprint recognition system by 50% and 40%, respectively. On the other hand, fingerprint matching task itself can be solved in multiple ways. Fingerprint feature extraction is a core step in the fingerprint matching task and it can be solved by using a global as well as a local approach. State of the art global approaches use huge deep learning models to process the full fingerprint image at once which makes the corresponding approach memory intensive. Whereas local approaches involve minutiae based patch extraction and multiple feature extraction steps which make the corresponding approach time-intensive. But both these approaches provide useful and sometimes exclusive insights for solving the problem. Hence using both kinds of approaches together for extracting fingerprint representations is semantically useful but quite inefficient. Hence, in the second study, we use a convolutional transformer based approach with an in-built minutiae extractor. This study provides an efficient solution for feature extraction that parallely provides a global as well as a local represen tation of a fingerprint. This not only provides a replacement for using a heavy combination of two independent local and global approaches but also provides a further 54.41% speedup over the local approach and a 57.93% conservation in memory as compared to the global approach. Both of these studies work towards the common goal of optimizing fingerprint recognition systems and hence provide a significant advantage in the context of such systems deployed on resource constrained devices like smartphones.

Abstract

Recent work in NLP has focused on applying powerful neural sequence models to various learning problems. These neural models excel at extracting statistical patterns from large amounts of data but struggle to learn patterns or reason about language from only a few examples. We ask the question: Can we learn explicit rules that generalize well from only a few examples? We explore this question by viewing linguistic rules as programs that operate on linguistic forms. This allows us to tackle the problem of learning linguistic rules using program synthesis – a method to learn rules in the form of programs in a domain-specific language (DSL). We develop a synthesis model to learn phonological rules as programs in a DSL. In addition to being highly sample-efficient, our approach generates human-readable programs and allows control over the generalizability of the learned programs. We test the ability of our models to generalise from only a few training examples using our new dataset of problems from the Linguistics Olympiad. These problems are tasks from contests for high school students around the world that require inferring linguistic patterns from a small number of given examples. These problems are a challenging set of tasks that require strong linguistic reasoning ability. Having shown that program synthesis can be used to learn phonological rules in highly data-constrained settings, we use the problem of phonological stress placement as a case to study how the design of the domain-specific language influences the generalisation ability when using the same learning algorithm. We find that encoding the distinction between consonants and vowels results in much better performance, and providing syllable-level information further improves generalization. Program synthesis, thus, provides a way to investigate how access to explicit linguistic information influences what can be learned from a small number of examples.

Abstract

Virtual Labs are a collection of college-level online experiments in the science and engineering disciplines. Virtual Labs provides a remote experimentation platform for hands-on learning. It is essential to analyze and improve the web performance of the Virtual Labs platform to enhance the user experience as well as to cope with the increased usage. This study identifies and implements approaches to improve the performance of Virtual Labs experiments. We gather the web performance data using two tools, Google Analytics and Webpagetest, analyze and implement the performance improvement approaches. The implemented approaches have reduced the load time of the Virtual Labs experiments by 80% and reduced the load time of the page’s main content from 6.1 seconds to under 1.5 seconds. The loading and rendering metrics have been optimized, and the efficiency of the experiment loading has increased.

Abstract

Music, an integral part of our lives, which is not only a source of entertainment but plays an important role in mental well-being by impacting moods, emotions and other affective states. In this age of Big Data, online music streaming services allow us to capture ecologically valid music listening behavior and provide a rich source of information to identify several user-specific aspects. Music preferences and listening strategies have been shown to be associated with the psychological well-being of listeners including internalized symptomatology and depression. However, till date no studies exist that examine music-related social tags, static & time-varying audio-derived features, in terms of acoustic and emotional content, and its association with users well-being. The current work aims at unearthing static and dynamic patterns and trends in the individuals at risk for depression as it manifests in naturally occurring online music consumption. Mental well-being scores and listening histories of 541 Last.fm users were examined. Social tags associated with each listener’s most popular tracks were analyzed to unearth the mood/emotions and genres associated with the users. Results revealed that social tags prevalent in the users at risk for depression were predominantly related to emotions depicting Sadness associated with genre tags representing neo-psychedelic-, avant garde-, dream-pop. Static and dynamic acoustic and emotion-related features were extracted from each user’s listening history and correlated with their mental well-being scores. Results revealed that individuals with greater depression risk resort to higher dependency on music with greater repetitiveness in their listening activity. Furthermore, the affinity of depressed individuals towards music that can be perceived as sad was found to be resistant to change over time. This study has large implications for future work in the area of assessing mental illness risk by exploiting digital footprints of users via online music streaming platforms. This study will open up avenues for an MIR-based approach to characterizing and predicting risk for depression which can be helpful in early detection and additionally provide bases for designing music recommendations accordingly

Abstract

Popularity prediction is a well-studied machine learning task with wide-ranging business applications. The primary goal in a popularity prediction problem is to estimate the future popularity of a piece of content that is put up on the Internet. For example, a popularity prediction model might estimate the number of page views that a particular online news article would accrue over a time period. Or a social-media popularity prediction system might try to foretell the number of comments or shares that some social media post would receive over a week. Being able to predict the prospective popularity of content within a very short time after its publication has multiple downstream applications in article recommendation and advertising, and document retrieval. Additionally, proactively forecasting document popularity effectively allows an editor to improve a document in order to make it more engaging. Though there has been much work on the document-level prediction of prospective popularity labels for news articles, online petitions, social media posts, etc., there have been no attempts to forecast information popularity at a more granular level. Sentence-level popularity forecasting is of interest not only because of the insights it produces about how informative sentences contribute to a document’s overall popularity but also because of its potential applications in tasks such as popularity-guided text summarization, pull quote selection, catchphrase extraction, webpage search engine optimization, etc. In this thesis, we overcome existing challenges to capture information popularity at the level of individual sentences, and we present neural models that effectively forecast sentence-specific information popularity. Formally, we introduce the task of proactively forecasting relative popularities of sentences within online news documents solely utilizing their natural language content. Instead of taking a simplistic route where one would predict binary popular-or-not labels for individual sentences, we frame sentencepopularity forecasting as a regression task where we assign a normalized score ∈ [0, 1] to each sentence indicating its intra-document information popularity relative to all other sentences. To overcome the absence of sentence-level labeled data for information popularity, we curate INFOPOP, a dataset containing popularity labels for over 1.7 million sentences from over 50,000 online news documents sourced from 26 reputed news websites. To the best of our knowledge, INFOPOP is the first dataset automatically created using streams of incoming search engine queries to generate sentence-level popularity annotations. We observe how certain aspects of information popularity are related to sentence salience and propose auxiliary transfer learning subtasks of sentence salience prediction for pretraining our neural sentence popularity forecasting models. Sentence salience is well studied in the context of automatic text summarization. In this thesis, we provide an overview of document summarization approaches and present our findings from two scholarly document summarization tasks. In particular, we describe an abstractive and an extractive approach for summarizing scientific research papers, which have led to state-of-the-art results on two shared tasks, namely LaySumm and LongSumm, at the Scholarly Document Processing workshop (SDP) at EMNLP ’20. We further discuss the distinction between salience and sentences’ summary-inclusion worthiness in an extractive summarization setting. For sentence-specific popularity forecasting, we experiment with unsupervised baselines, a rudimentary recurrent neural network-based model, and a robust sentence-sequence regression architecture based on BERT (Bidirectional Encoder Representations from Transformers). Our results show that our proposed Transfer Learning approach produces significant performance gains on all metrics – reaching nDCG values over 0.8. We conduct extensive evaluations to highlight insightful distinctions between our primary and auxiliary tasks. Notably, our findings showcase a non-trivial takeaway: though information popularity and information salience are varying concepts, transfer learning from salience prediction significantly enhances sentence popularity forecasting.

Abstract

Dementia is a neurological disease that affects people starting around the age of 45. The most common symptom of dementia includes memory difficulties, impaired judgment-making capacity, difficulty in performing daily living activities, mind wandering, and disorientation (WHO 2022). Researchers have been looking into the causes, rate of progression, and potential remedies for many years. However, there is as yet no cure for dementia. As a result, researchers have been investigating the efficacy of alternative therapeutic methods such as regular exercise, therapies, meditation, etc. The present thesis has focused on investigating dementia Alzheimer's type and its early phases. It includes an overview of Alzheimer's disease and potential research gaps related to atrophy rates in selected regions of interest, such as the hippocampal subfields and the effects of meditation as adjuvant therapy. We used longitudinal structural and functional MRI datasets from humans to address these two research topics in the thesis. We have explored the nature of atrophy in the hippocampal subfields during the distinct phases of Alzheimer's disease. This research hypothesized that hippocampal subfields in Alzheimer's disease patients would have the highest gray matter volume atrophy rate compared to other groups (control, early MCI, and late MCI). We have also proposed that the subfields being involved in memory encoding and decoding would deteriorate faster in the late MCI or AD groups. The analysis used structural MRI data available at two-time points (separated by a year) from a cohort of 196 participants in the publicly available ADNI dataset. We have estimated the gray matter volume (as the cortical thickness is irrelevant here in the subcortical region) and compared it across the four groups. Subfields that are mainly involved in the process of memory encoding, consolidation, and decoding, such as the dentate gyrus, CA3, CA1, and subiculum, have shown a higher atrophy rate in Alzheimer's disease and late MCI groups as compared to early MCI or healthy controls over a one-year time period.We have also investigated the long-term effects of meditation techniques (one of the alternative therapy approaches) in early MCI or Alzheimer's disease patients over the course of six months. This study hypothesized that simple long-term daily meditation would have a facilitatory effect on the brain regions related to attention and executive control in patients with early MCI or Alzheimer's disease. In his study, we have used a dataset collected by our collaborators on a cohort of dementia patients from Kolkata, India. In this longitudinal study, we compared the meditation and non-meditation groups on two structural brain measures: cortical thickness (CT) and gray matter volume (GMV). Our results indicate that long-term daily meditation led to substantial improvements in cortical thickness and gray matter volume in the regions responsible for attention and executive control in the prefrontal and inferior parietal cortex. In summary, the thesis presents results from two longitudinal studies. One study characterized the atrophy rate differences among aged controls and early-to-late-stage dementia patients. The other study investigated the efficacy of meditation in arresting the structural decline in some critical brain regions. It is hoped that these studies are useful for the clinical management of dementia.

Abstract

Machine learning techniques have found widespread usage in critical systems in the recent past.This rise of ML was mainly fueled by Neural Networks, which enabled end-to-end systems to use.This is possible because of the feature learning and universal approximator properties of NNs. Recentresearch has extended the application of NNs to the estimation of mutual information (MI) from samples.They construct a critic using a NN and maximize a variation lower bound of MI. These methods havealleviated many problems faced by previous MI estimation methods like the inability to scale andincompatibility with mini-batch-based optimization. While these estimation methods are reliable whenthe true MI is low, they tend to produce high variance estimates when the true MI is high. We arguethat the high variance characteristics are due to the uncontrolled complexity of the critic’s hypothesisspace. In support of this argument, we use the data-drivenRademacher complexityof the hypothesisspace associated with the critic’s architecture to derive upper bounds on thegeneralization errorofvariational lower bound based estimates of MI. Using the derived bounds, we show that Rademachercomplexity of the critic’s hypothesis is a major contributor to the generalization error. Inspired by thiswe propose to negate the high variance characteristics of these estimators by constraining the critic’shypothesis space to aReproducing Kernel Hilbert Space(RKHS), which corresponds to a kernel learnedusingAutomated Spectral Kernel Learning(ASKL). Further, we propose two regularisation terms byexamining the upper bound on the Rademacher complexity of RKHS learned by ASKL. The weightsof these regularization terms can be varied for a bias-variance tradeoff. We empirically demonstratethe efficacy of this regularization in enforcing proper bias-variance tradeoff on four different variationallower bounds of MI, namely NWJ, MINE, JS and SMILE.We also propose a novel application of MI estimation using variational lower bounds and NNs inlearning disentangled representations of videos. The learned representations contain two sets of features,content and pose representations which disjointly represent the content and movement in the video. Thisis achieved by minimizing the mutual information between pose representation of different frames from the same video, thus in effect forcing the content representation to capture information common betweendifferent frames. We qualitatively and quantitatively compare our method against DRNET [24] on twosynthetic and one real-world dataset.