Jyoti D. Pawar

Bio: Jyoti D. Pawar is an academic researcher from Goa University. The author has contributed to research in topics: Konkani & Dance. The author has an hindex of 8, co-authored 65 publications receiving 236 citations.

The WordNet in Indian Languages

Abstract: This contributed volume discusses in detail the process of construction of a WordNet of 18 Indian languages, called Indradhanush (rainbow) in Hindi. It delves into the major challenges involved in developing a WordNet in a multilingual country like India, where the information spread across the languages needs utmost care in processing, synchronization and representation. The project has emerged from the need of millions of people to have access to relevant content in their native languages, and it provides a common interface for information sharing and reuse across the Indian languages. The chapters discuss important methods and strategies of language computation, language data processing, lexical selection and management, and language-specific synset collection and representation, which are of utmost value for the development of a WordNet in any language. The volume overall gives a clear picture of how WordNet is developed in Indian languages and how this can be utilized in similar projects for other languages. It includes illustrations, tables, flowcharts, and diagrams for easy comprehension. This volume is of interest to researchers working in the areas of language processing, machine translation, word sense disambiguation, culture studies, language corpus generation, language teaching, dictionary compilation, lexicographic queries, cross-lingual knowledge sharing, e-governance, and many other areas of linguistics and language technology.

Non-invasive hyperspectral imaging approach for fruit quality control application and classification: case study of apple, chikoo, guava fruits

Abstract: Mechanical injuries to fruits are often caused due to hidden internal damages that results in bruising of fruit. This is a serious cause of concern to the fruit industry, as spoiled or bruised fruits directly impact the producers profit. Hyperspectral imaging method can provide the ability to identify these internal bruises to classify these fruits as normal and injured (bruised), reducing time and increasing efficiency over the sorting line in marketing chain. In this paper, we have used three types of fruits i.e., apple, chikoo & guava for experiments. The mechanical injury is introduced by manual impact on surface of the fruits sample and hyperspectral images were captured over nine narrow band pass filters to produce hyperspectral cubes for a fruit. Three types of methods were used for the data processing. First two are non-invasive in nature i.e., pixel signatures over hyperspectral cubes and second is prediction model for classification of fruits quality into normal and bruised using feed forward back propagation neural network. Finally, invasive method is used to confirm the said prediction model using parameters like firmness, Total Soluble Solid (TSS) and weight with Principal Component Analysis. Results obtained by hyperspectral imaging method indicate scope for non-invasive quality control over spectral wavelength range of 400–1000 nm.

Use of an Evolutionary Approach for Question Paper Template Generation

Abstract: This paper focuses on question paper template generation and its use in dynamic generation of examination question paper. Question paper template generation is a constrained based optimization problem. Choosing an efficient, scientific and rational algorithm to generate a template is the key to dynamic examination question paper generation. By using the evolutionary computational search technique of evolutionary programming and educational taxonomies, this paper analyses and experimentally proves that the generated question paper templates are best suited for dynamic examination paper generation. This new technique outperforms traditional algorithms in terms of coverage of topics, learning domains and marks distribution in the generated question paper.

Art to SMart: An evolutionary computational model for BharataNatyam choreography

Abstract: Dance choreography is an intense, creative and intuitive process. A choreographer has to finalize appropriate dance steps from amongst millions of possibilities. Though it is not impossible, the choreographer being human cannot explore, analyze and remember all these variations among steps due to large scale of available options. Hence, we propose to simplify the problem of exploring and selecting dance steps from amongst the huge set of all possible variations for an Indian Classical Dance, BharataNatyam (BN). Based on a computational model developed by Jadhav et al. [13], we propose a Genetic Algorithm (GA) driven automatic system that would provide a list of unexplored novel dance steps to choreographers. We have incorporated certain measures to ensure that the proposed dance steps should be feasible and appropriate. In this paper, we discuss an automated approach to obtain unexplored dance steps using a proposed fitness function for a single beat/count. The details of experimental study performed for the Genetic Algorithm based art to SMart (System Modelled art) system along with the results obtained are also presented in this paper.

Modeling BharataNatyam dance steps: art to SMart

Abstract: BharataNatyam (BN) like any other Indian classical dance comprises of a sequence of possible and legitimate dance steps. It is estimated that using the main body parts namely head, neck, hand and leg itself, more than 5 lakh dance steps can be generated for a single beat. Choreographers and even dancers usually repeat their favorite dance steps or the conventional casual dance steps taught by their teacher while performing for multiple beats. As a result several valid and many other significant non-traditional dance steps remain unexplored. Hence, we propose to have an auto enumeration followed by auto classification of significant BN dance steps that can be used in dance performance and choreography. In short, we try to transform sheer art into a System Modelled art i.e. 'Art to SMart'. The foremost and most challenging task is to have a computational model that represents different BN dance poses. In this paper, we have proposed a computational model to represent BN dance steps and have presented the detailed description of formulation of a dance position vector that comprises of thirty explicitly identified attributes to capture and represent all variations of a BN dance step. We have named it as a SMart system for modelling BN steps, where SMart stands for System Modelled art. We have also demonstrated sample dance steps and their corresponding representations with appropriate dance step images.

Document Analysis and Recognition

Abstract: The subject about document image understanding is to extract and classify individual data meaningfully from paper-based documents. Until today, many methods/approaches have been proposed with regard to recognition of various kinds of documents, various technical problems for extensions of OCR, and requirements for practical usages. Of course, though the technical research issues in the early stage are looked upon as complementary attacks for the traditional OCR which is dependent on character recognition techniques, the application ranges or related issues are widely investigated or should be established progressively. This paper addresses current topics about document image understanding from a technical point of view as a survey. key words: document model, top-down, bottom-up, layout structure, logical structure, document types, layout recognition

How Intense Are You? Predicting Intensities of Emotions and Sentiments using Stacked Ensemble [Application Notes]

Abstract: Emotions and sentiments are subjective in nature. They differ on a case-to-case basis. However, predicting only the emotion and sentiment does not always convey complete information. The degree or level of emotions and sentiments often plays a crucial role in understanding the exact feeling within a single class (e.g., `good' versus `awesome'). In this paper, we propose a stacked ensemble method for predicting the degree of intensity for emotion and sentiment by combining the outputs obtained from several deep learning and classical feature-based models using a multi-layer perceptron network. We develop three deep learning models based on convolutional neural network, long short-term memory and gated recurrent unit and one classical supervised model based on support vector regression. We evaluate our proposed technique for two problems, i.e., emotion analysis in the generic domain and sentiment analysis in the financial domain. The proposed model shows impressive results for both the problems. Comparisons show that our proposed model achieves improved performance over the existing state-of-the-art systems.

Predicting academic performance: a systematic literature review

Abstract: The ability to predict student performance in a course or program creates opportunities to improve educational outcomes. With effective performance prediction approaches, instructors can allocate resources and instruction more accurately. Research in this area seeks to identify features that can be used to make predictions, to identify algorithms that can improve predictions, and to quantify aspects of student performance. Moreover, research in predicting student performance seeks to determine interrelated features and to identify the underlying reasons why certain features work better than others. This working group report presents a systematic literature review of work in the area of predicting student performance. Our analysis shows a clearly increasing amount of research in this area, as well as an increasing variety of techniques used. At the same time, the review uncovered a number of issues with research quality that drives a need for the community to provide more detailed reporting of methods and results and to increase efforts to validate and replicate work.

A multi-scale deep quad tree based feature extraction method for the recognition of isolated handwritten characters of popular indic scripts

Abstract: Recognition of handwritten characters is a challenging task Variations in writing styles from one person to another, as well as for a single individual from time to time, make this task harder Hence, identifying the local invariant patterns of a handwritten character or digit is very difficult These challenges can be overcome by exploiting various script specific characteristics and training the OCR system based on these special traits Finding ubiquitous invariant patterns and peculiarities, applicable for handwritten characters or digits of multiple scripts, is much more difficult In the present work, a non-explicit feature based approach, more specifically, a multi-column multi-scale convolutional neural network (MMCNN) based architecture has been proposed for this purpose A deep quad-tree based staggered prediction model has been proposed for faster character recognition These denote the most significant contributions of the present work The proposed methodology has been tested on 9 publicly available datasets of isolated handwritten characters or digits of Indic scripts Promising results have been achieved by the proposed system for all of the datasets A comparative analysis has also been performed against some of the contemporary OCR systems to prove the superiority of the proposed system We have also evaluated our system on MNIST dataset and achieved a maximum recognition accuracy of 9974%, without any data augmentation to the original dataset