Search

Chits funds are typically operated among a trusted network of people, by a foreman, completely relying on the trust and transparency of the transactional data audited through the network. The cost involved in this ledger keeping or audit procedures for ensuring trust and transparency is the major institutional overhead for running a chit fund. This paper proposes a transparent platform based on distributed ledger technology termed Blockchain as a Service (BaaS) for setting up the ChitChain infrastructure. The service providers and the end-users (customers) will be provided with client interfaces for the secured access methods to the system. As the notion of trust will be taken from the institution to the tamper-proof platform, the approach discussed in this paper will propose a decentralized finance model that enables higher financial inclusion. The auditable platform provides insights to the stakeholders regarding the transactions and operations of the chit fund. This work proposes a novel operation model for setting up a chitty, a fair jury selection protocol for decentralized management, and an efficient foreman and end-user onboarding process.

Online social networks are considered to be one of the most disruptive platforms where people communicate with each other on any topic ranging from funny cat videos to cancer support. The widespread diffusion of mobile platforms such as smart-phones causes the number of messages shared in such platforms to grow heavily, thus more intelligent and scalable algorithms are needed for efficient extraction of useful information. This paper proposes a method for retrieving relevant information from social network messages using a distributional semantics-based framework powered by topic modeling. The proposed framework combines the Latent Dirichlet Allocation and distributional representation of phrases (Phrase2Vec) for effective information retrieval from online social networks. Extensive and systematic experiments on messages collected from Twitter (tweets) show this approach outperforms some state-of-the-art approaches in terms of precision and accuracy and better information retrieval is possible using the proposed method.

The emergence of deep learning frameworks paves the way for achieving higher-level data abstractions and possess the potential in consolidating both supervised and unsupervised learning paradigms. Researchers have made many successful explorations in the field of deep learning, with applications in the fields of face recognition, text mining, language translation, image prediction, and action recognition. Kernel machines act as a bridge between the linearity and nonlinearity for many machine learning algorithms such as support vector machines, extreme learning machines, and core vector machines. These Kernel machines play a vital role in mapping the data in the input space to a Kernel-induced high-dimensional feature space to obtain a better distribution of the data. In this Kernel-induced high-dimensional feature space, the distribution of data points will be more amenable to the classification problem under consideration. The Kernel trick facilitates in transforming the machine learning algorithms that require only inner product computations between the data vectors into a Kernel-based approach by selecting an appropriate Kernel function. In Kernel-based approaches, the Kernel functions can thus be utilized for accomplishing the inner product computations between the transformed data vectors in an implicitly defined Kernel-induced feature space. Unlike neural networks, the Kernel machines guarantee structural risk minimization and global optimal solutions. Also, the Kernel machines exhibit capabilities such as theoretical tractability and excellent performance in practical applications. These attempts motivated the researchers towards utilizing the emerging trends of deep learning with Kernel methods for building deep Kernel machines. Researchers integrate Kernel methods and deep learning networks for maintaining their advantages and make up their limitations, then apply the deep Kernel learning approaches for improving the performance of the learning algorithm in different applications. Different ways of building deep Kernel machines by integrating the Kernel methods and deep learning architectures include utilizing Kernel machines as the final classifier of deep learning networks, Kernelization in deep neural networks for better feature enrichment, and building deep Kernel machines by utilizing deep or multiple Kernels in different tasks. This survey attempts to provide an overview of different approaches in building several deep Kernel learning architectures for enhancing the learning algorithm properties and their performance in practical applications.

Advances in information technology are digitizing the healthcare domain with the aim of improved medical services, diagnostics, continuous monitoring using wearables, etc., at reduced costs. This digitization improves the ease of computation, storage and access of medical records which enables better treatment experiences for patients. However, it comes with a risk of cyber attacks and security and privacy concerns on this digital data. In this work, we propose a Blockchain based solution for healthcare records to address the security and privacy concerns which are currently not present in existing e-Health systems. This work also explores the potential of building trusted Artificial Intelligence models over Blockchain in e-Health, where a transparent platform for consent-based data sharing is designed. Provenance of the consent of individuals and traceability of data sources used for building and training the AI model is captured in an immutable distributed data store. The audit trail of the data access captured using Blockchain provides the data owner to understand the exposure of the data. It also helps the user to understand the revenue models that could be built on top of this framework for commercial data sharing to build trusted AI models.

Emergence of extreme learning machine as a breakneck learning algorithm has marked its prominence in solitary hidden layer feed-forward networks. Kernel-based extreme learning machine (KELM) reflected its efficiency in diverse applications where feature mapping functions of hidden nodes are concealed from users. The conventional KELM algorithms involve only solitary layer of kernels, thereby emulating shallow learning architectures for its feature transformation. Trend in migrating shallow-based learning models into deep learning architectures opens up a new outlook for machine learning domains. This paper attempts to bestow deep kernel learning approach in a conventional shallow architecture. The emerging arc-cosine kernels possess the potential to mimic the prevailing deep layered frameworks to a greater extent. Unlike other kernels such as linear, polynomial and Gaussian, arc-cosine kernels have a recursive nature by itself and have the potential to express multilayer computation in learning models. This paper explores the possibility of building a new deep kernel machine with extreme learning machine and multilayer arc-cosine kernels. This framework outperforms conventional KELM and deep support vector machine in terms of training time and accuracy.

Support Vector Clustering (SVC) algorithm reformulates SVM’s Quadratic Programming as a minimum enclosing ball (MEB) problem, where every point in the data space will be projected to the higher dimensional feature space to find the minimum radius that encloses all the data points inside the sphere. Support vectors are data points in the surface of the minimum enclosing ball. The clustering algorithm works by mapping the support vectors of MEB back to the data plane forming groups or derives a contour that encloses a set of clustered points. However, the major limitation of this algorithm is that it fails to scale with larger dataset. Computation bottleneck lies in the efficient cluster computation approach and in solving the QP optimization for MEB. In order to solve this cluster computational complexity, this work presents a simple, efficient cluster assignment algorithm using similarity of feature set for data points in high-dimensional feature space utilizing an efficient MEB approximation algorithm. Experiments are carried out by varying the similarity threshold metrics. Performance of the clustering mechanism run on various datasets demonstrates the proposed algorithm is lighter, simple, and converges faster.

The developments in information and communication technologies pave the way to the digitization of traditional handwritten records to the computerized methods of record creation and storage. The healthcare industry is no different in adopting digitization mechanisms and the rate of adoption to the electronic health records (EHR) is rapid. With the introduction of many acts and regulations such as the Health Information Technology for Economic and Clinical Health Act introduced in 2009 and associated regulations such as Health information exchange act and False Claim Act, the rate has become exponential. It is evident that EHR hold beneficial knowledge for healthcare researchers and practitioners, and thus it is desirable to extract interesting patterns from the text. Clinical decision support systems heavily make use of information extracted from unstructured EHR for decision-making purposes. Clinical research studies also use such patterns extracted from unstructured EHR for analysis and predictions. Information extraction generally deals with automatic extraction of keyphrases, concepts, events, entities, and their relationships from unstructured text. This is a highly empirical area in natural language processing that uses lexical, syntactic, and semantic features for the extraction of useful information. As the volume and variety of the data increase, traditional rule-based information extraction systems will not perform well on such data and thus we need better algorithms that operate on text taking into consideration the meaning or semantics. A meaning-aware text processing algorithm can better extract patterns from highly unstructured text and can aid in the process of creating ontologies which are considered to be the backbone of semantic computing systems. Therefore it is evident that the research and development of technological breakthroughs in the areas of meaning-aware information extraction from healthcare documents and other electronic records is a highly demanding opportunity. This chapter is a novel attempt in this dimension that discusses a framework for building a meaning-aware healthcare information extraction from unstructured EHR. The proposed framework uses medical ontologies, medical catalog-based terminology extractor, and a semantic reasoner to build the medical knowledge base that is used for enabling a semantic information search and retrieval experience in the healthcare domain. An illustration of the same is also included in this chapter to better explain the proposed framework along with some future research dimensions on enhancing the framework discussed in this chapter. The empirical studies conducted show that the proposed framework outperforms some of the information extraction systems already available in the literature.

Blockchain is a distributed data store that offers a trusted infrastructure for mutually distrusting parties to network and transact, eliminating the need of any central authority or third party intermediary. The trust element in blockchain is due to the tamper proof record keeping in blockchain through efficient hashed block-linking mechanism and the consensus protocol among the network peers to create any new record in the distributed ledger. The consensus mechanism in distributed data store ensures that every participant node in the network accepts on a common agreement over a transaction or state transition in a system by gathering confirmation from all or at-least majority of the members. Usually some form of voting or lottery mechanism is employed for the leader election process to implement consensus among the participants in a business network. Consensus mechanism is also responsible for the smooth functioning of the distributed network in the event of some arbitrary faults or node failures or even any malicious attacks in the network. Generally a consensus mechanism implementation is a tailored solution based on application or use case. It ensures the correctness of data in the trusted distributed ledger through the peer agreement process. In this work, we explore the evolution of state of the art consensus protocols for various blockchain implementations, their application, advantages and drawbacks as well.

Over the last few years, we have been witnessing a dramatic progress of deep learning in many real world applications. Deep learning concepts have been originated in the area of neural network and show a quantum leap in effective feature learning techniques such as auto-encoders, convolutional neural networks, recurrent neural networks etc. In the case of kernel machines, there are several attempts to model learning machines that mimic deep neural networks. In this direction, Multilayer Kernel Machines (MKMs) was an attempt to build a kernel machine architecture with multiple layers of feature extraction. It composed of many layers of kernel PCA based feature extraction units with support vector machine having arc-cosine kernel as the final classifier. The other approaches like Multiple Kernel Learning (MKL) and deep core vector machines solve the fixed kernel computation problem and scalability aspects of MKMs respectively. In addition to this, there are lot of avenues where the use of unsupervised MKL with both single and multilayer kernels in the multilayer feature extraction framework have to be evaluated. In this context, this paper attempts to build a scalable deep kernel machines with multiple layers of feature extraction. Each kernel PCA based feature extraction layer in this framework is modeled by the combination of both single and multilayer kernels in an unsupervised manner. Core vector machine with arc-cosine kernel is used as the final layer classifier which ensure the scalability in this model. The major contribution of this paper is a novel effort to build a deep structured kernel machine architecture similar to deep neural network architecture for classification. It opens up an extendable research avenue for researchers in deep learning based intelligent system leveraging the principles of kernel theory. Experiments show that the proposed method consistently improves the generalization performances of existing deep core vector machine.

Conventional Extreme Learning Machines utilize Moore–Penrose generalized pseudo-inverse to solve hidden layer activation matrix and perform analytical determination of output weights. Scalability is the major concern to be addressed in Extreme Learning Machines while dealing with large dataset. Motivated by these scalability concerns, this paper proposes a novel tensor decomposition based Extreme Learning Machine which utilize PARAFAC and TUCKER decomposition based techniques in a SPARK platform. This proposed Extreme Learning Machine achieve reduced training time and better accuracy when compared with a conventional Extreme Learning Machine.

In machine learning literature, deep learning methods have been moving toward greater heights by giving due importance in both data representation and classification methods. The recently developed multilayered arc-cosine kernel leverages the possibilities of extending deep learning features into the kernel machines. Even though this kernel has been widely used in conjunction with support vector machines (SVM) on small-size datasets, it does not seem to be a feasible solution for the modern real-world applications that involve very large size datasets. There are lot of avenues where the scalability aspects of deep kernel machines in handling large dataset need to be evaluated. In machine learning literature, core vector machine (CVM) is being used as a scaling up mechanism for traditional SVMs. In CVM, the quadratic programming problem involved in SVM is reformulated as an equivalent minimum enclosing ball problem and then solved by using a subset of training sample (Core Set) obtained by a faster $$(1+\epsilon )$$ approximation algorithm. This paper explores the possibilities of using principles of core vector machines as a scaling up mechanism for deep support vector machine with arc-cosine kernel. Experiments on different datasets show that the proposed system gives high classification accuracy with reasonable training time compared to traditional core vector machines, deep support vector machines with arc-cosine kernel and deep convolutional neural network.

In the original publication of the article, the Acknowledgements section was missed unfortunately. The correct Acknowledgements section is given in this correction.

Artificial Intelligence Marketingis a form of direct marketing which leverages the techniques of database marketing along with the concepts and models of AI such as machine learning. Direct marketi...

This paper proposes an efficient and scalable method for concept extraction and concept hierarchy learning from large unstructured text corpus which is guided by a topic modeling process. The method leverages “concepts” from statistically discovered “topics” and then learns a hierarchy of those concepts by exploiting a subsumption relation between them. Advantage of the proposed method is that the entire process falls under the unsupervised learning paradigm thus the use of a domain specific training corpus can be eliminated. Given a massive collection of text documents, the method maps topics to concepts by some lightweight statistical and linguistic processes and then probabilistically learns the subsumption hierarchy. Extensive experiments with large text corpora such as BBC News dataset and Reuters News corpus shows that our proposed method outperforms some of the existing methods for concept extraction and efficient concept hierarchy learning is possible if the overall task is guided by a topic modeling process.

Distributional Semantics is an active research area in natural language processing (NLP) that develop methods for quantifying semantic similarities between linguistic elements in large samples of data. Short text conceptualization on the other hand is a technique for enriching short texts so that it become more interpretable. This is needed because most text mining tasks including topic modeling and clustering are based on statistical methods and won’t consider the semantics of text. This paper proposes a novel framework for combining distributional semantics and short text conceptualization for better interpretability of phrases in text data. Experiments on real-world datasets show that this method can better enrich phrases that are represented in distributional semantic spaces.

This paper proposes a framework which induces semantically rich concepts from probabilistically generated topics by a topic modeling algorithm. In this method an off-the-shelf tool has been used to extract noun-phrases as word bi-grams and tri-grams from the static document corpus and then models the topics using Latent Dirichlet Allocation algorithm. Additionally, we show that a small extension to our proposed framework can better rank documents in a large collection, which is a well studied area in information retrieval. Experiments conducted on three real world datasets show that this proposed framework outperforms state-of-the-art methods used for extracting concepts and ranking documents. When compared with the baselines chosen, our proposed concept extraction method showed an increased f-measure in the range of 16.65% to 22.04% and the proposed topic modeling guided document retrieval method showed 7.6%–16.61% increase in f-measure.

The task of mining large unstructured text archives, extracting useful patterns and then organizing them into a knowledgebase has attained a great attention due to its vast array of immediate applications in business. Businesses thus demand new and efficient algorithms for leveraging potentially useful patterns from heterogeneous data sources that produce huge volumes of unstructured data. Due to the ability to bring out hidden themes from large text repositories, topic modeling algorithms attained significant attention in the recent past. This paper proposes an efficient and scalable method which is guided by topic modeling for extracting concepts and relationships from e-commerce product descriptions and organizing them into knowledgebase. Semantic graphs can be generated from such a knowledgebase on which meaning aware product discovery experience can be built for potential buyers. Extensive experiments using proposed unsupervised algorithms with e-commerce product descriptions collected from open web shows that our proposed method outperforms some of the existing methods of leveraging concepts and relationships so that efficient knowledgebase construction is possible.

Maximisation of influence propagation is a key ingredient to any viral marketing or socio-political campaigns. However, it is an NP-hard problem, and various approximate algorithms have been suggested to address the issue, though not largely successful. In this paper, we propose a bio-inspired approach to select the initial set of nodes which is significant in rapid convergence towards a sub-optimal solution in minimal runtime. The performance of the algorithm is evaluated using the re-tweet network of the hashtag #KissofLove on Twitter associated with the non-violent protest against the moral policing spread to many parts of India. Comparison with existing centrality based node ranking process the proposed method significant improvement on influence propagation. The proposed algorithm is one of the hardly few bio-inspired algorithms in network theory. We also report the results of the exploratory analysis of the network kiss of love campaign.

Research in social networks is attaining more attention in the recent past due to the explosive growth in the creation and sharing of information over social media. As the volume of information grows exponentially, we need efficient computational techniques to analyze this information and to synthesis the hidden knowledge associated with it. Being a suit of text understanding algorithms, topic modeling discovers the topics or themes within a huge collection of documents. In this work, we employ the essence of a powerful topic modeling algorithm to analyze hidden knowledge contained in the information spread across a famous social network platform Twitter, using a novel iterative topic modeling approach. Additionally, we visualized the knowledge extracted using a sunburst chart so that even a naive user can interpret the hidden knowledge extracted from tweets.

Cluster analysis has been widely applied in many areas such as data mining, geographical data processing, medicine, classification of statistical findings in social studies and so on. Most ofthese domains deal with massive collections of data. Hence the methods to handle them must be efficient both in terms of the number of data set scans and memory usage.

Even though several techniques have been proposed in the literature for achieving multiclass classification using Support Vector Machine(SVM), the scalability aspect of these approaches to handle large data sets still needs much of exploration. Core Vector Machine(CVM) is a technique for scaling up a two class SVM to handle large data sets. In this paper we propose a Multiclass Core Vector Machine(MCVM). Here we formulate the multiclass SVM problem as a Quadratic Programming(QP) problem defining an SVM with vector valued output. This QP problem is then solved using the CVM technique to achieve scalability to handle large data sets. Experiments done with several large synthetic and real world data sets show that the proposed MCVM technique gives good generalization performance as that of SVM at a much lesser computational expense. Further, it is observed that MCVM scales well with the size of the data set.

Support vector machines (SVMs) are hyperplane classifiers defined in a kernel induced feature space. The data size dependent training time complexity of SVMs usually prohibits its use in applications involving more than a few thousands of data points. In this paper, we propose a novel kernel based incremental data clustering approach and its use for scaling non-linear support vector machines to handle large data sets. The clustering method introduced can find cluster abstractions of the training data in a kernel induced feature space. These cluster abstractions are then used for selective sampling based training of support vector machines to reduce the training time without compromising the generalization performance. Experiments done with real world datasets show that this approach gives good generalization performance at reasonable computational expense

Core vector machine(CVM) is suitable for efficient large-scale pattern classification. In this paper, a method for improving the performance of CVM with Gaussian kernel function irrespective of the orderings of patterns belonging to different classes within the data set is proposed. This method employs a selective sampling based training of CVM using a novel kernel based scalable hierarchical clustering algorithm. Empirical studies made on synthetic and real world data sets show that the proposed strategy performs well on large data sets.

This paper discusses a method for scaling SVM with Gaussian kernel function to handle large data sets by using a selective sampling strategy for the training set. It employs a scalable hierarchical clustering algorithm to construct cluster indexing structures of the training data in the kernel induced feature space. These are then used for selective sampling of the training data for SVM to impart scalability to the training process. Empirical studies made on real world data sets show that the proposed strategy performs well on large data sets.

Semantic Web technologies leveraging ontological knowledge enable machine interpretable data to be published over the web. Modern day e-business initiatives are striving to utilise the potential of this emerging technology to influence and develop the internet economy. In this direction, web information discovery channels like search engines have attempted and partially succeeded in building rich semantic search experiences. These semantic search experiences are built over ontological knowledge bases which, in turn, are built by subject matter experts. In this paper, we discuss the scope of semantic annotation of product pages/catalogues in e-commerce websites for providing a better shopping experience. We discuss how a semantics driven product discovery mechanism and a meaning aware recommender system can be built over a semantically annotated product catalogue database.