School of Computer Sciences

The study was conducted in Palamu district of Jharkhand to know the socioeconomic status of farmers. They were categorized in three groups i.e. small, medium and large farmers which were differ from each other with regard to socioeconomic factors and found that 54.66 per cent rice growers has medium level of knowledge. Most of the small category farmers have high technological gap of lack of technological guidance (69.60%) followed by lack of money (66.13 %). The different sub practices grouped into 6 categories and found that plant protection measures is one of the major concerns for technological gap First spray of Carbendazim 50 WP @ 2 g/l of water before panicle initiation and second at milking stage Propiconazole 25EC @ 1 mi/l of water was effective in reducing incidence of false smut diseases and increased the grain yields.

As the research in sentiment analysis (SA) has reached the advanced level of maturity, scientific community is exploring possibilities of applying SA in different domains for which huge descriptive data is available. Researcher has developed tools for SA in commercial domains like feedback from customers, review rating of movie, political domains like exit poll prediction from tweets, etc. In medical science, huge data is available which can be used for predictive modelling. Psychomedical diagnosis is an interesting field which deals with exploring behavioural patterns to potential psychological problems. In the domain of medical science, little amount of work has been done in clinical SA. However, the efforts are at an elementary level in the course of research of SA for behavioural psychology. Hence, there is an opportunity of applying SA as the first step towards automatic psychomedical diagnosis of teenagers' behaviour patterns.

The objective of this research is to propose a dynamic composition of behaviour-rich interactive 3D scene as a virtual environment for cognitive support using visual and linguistic analytics. Based on the constructivist theory of learning through visual cognition by Winn (A conceptual basis for educational applications of virtual reality, Technical report TR 93-9, Washington Technology University, 1993), we propose our work with a grounding that visual data are easy to comprehend for a person having linguistic learning difficulties. Virtual reality provides an environment for learners to actively pursue their knowledge needs by applying their theories in the 'real world'. Therefore, we focus our work on generating an interactive virtual environment. It decreases cognitive load for the person with difficulties in comprehension, especially in language reading, e.g. dyslexia. Our prior work related to the proposed research is named as Preksha-an automatic Hindi text visualizer. To the best of the knowledge of the authors, Preksha is the only known visualization work for an Indian Language, viz. Hindi. Belonging to morphologically-rich and free-word order Indian languages, this work on the Hindi Language is a novel interdisciplinary approach to develop a virtual environment for cognitive support. Application of an automatic text visualization with a suitable learning paradigm in a virtual environment is another novelty of this research.

The study was conducted in Palamu district of Jharkhand to know the socioeconomic status of farmers. They were categorized in three groups i.e. small, medium and large farmers which were differ from each other with regard to socioeconomic factors and found that 54.66 per cent rice growers has medium level of knowledge. Most of the small category farmers have high technological gap of lack of technological guidance (69.60%) followed by lack of money (66.13 %). The different sub practices grouped into 6 categories and found that plant protection measures is one of the major concerns for technological gap First spray of Carbendazim 50 WP @ 2 g/l of water before panicle initiation and second at milking stage Propiconazole 25EC @ 1 mi/l of water was effective in reducing incidence of false smut diseases and increased the grain yields.

We have developed Preksha: a Hindi text visualizer. Preksha is an automatic text visualization [ATV] system. It generates a 3D scene by understanding a given Hindi text. This is the process of natural language understanding and conversion of this textual information into 3D scene generation. In this paper, we discuss the challenges for evaluation of ATV system. The area of text visualization is comparative new in discipline. It does not have much proven work toward the finalizing standards and strategies for Evaluation process. We present an evaluation plan for Preksha and initial results of Preksha-Evaluation. IndexTerms Artificial Intelligence, text-to-conversion, Natural Language Visualization, Human Computer Interface, Natural language processing, Hindi Language. _________________________________________________________________________________________________________________

Information Extraction (IE) is a sub discipline of Artificial Intelligence. IE identifies information in unstructured information source that adheres to predefined semantics i.e. people, location etc. Recognition of named entities (NEs) from computer readable natural language text is significant task of IE and natural language processing (NLP). Named entity (NE) extraction is important step for processing unstructured content. Unstructured data is computationally opaque. Computers require computationally transparent data for processing. IE adds meaning to raw data so that it can be easily processed by computers. There are various different approaches that are applied for extraction of entities from text. This paper elaborates need of NE recognition for Marathi and discusses issues and challenges involved in NE recognition tasks for Marathi language. It also explores various methods and techniques that are useful for creation of learning resources and lexicons that are important for extraction of NEs from natural language unstructured text.

Named Entity Recognition (NER) is sub task of Information Extraction that includes identification of named entities and classification of them into named entity classes such as person, location and organization etc. NER can be used to preprocess textual information and convert it into structured form that can be useful for Information Retrieval, Machine Translation, Question Answering System and Text Summarization. This paper presents a survey regarding NER research done for various Indian and non Indian languages. The study and observations related to approaches, techniques and features required to implement NER for various languages especially for Indian languages is reported.

This paper describes a named entity recognition system that combines hidden markov model, handcrafted rules, and gazetteers to recognize named entities in Marathi language. The objective of the system is to recognize twelve types of NEs from the Marathi text. Marathi is morphologically rich and inflectional language. The inflections in NEs are handled by using lemmatization. The difficulties of zero and poor probabilities caused due to the sparse data are handled using pseudo word replacement and smoothing techniques. Viterbi algorithm is used for decoding and word disambiguation. The performance of the system is improved using gazetteers and grammar rules.

The present manuscript focuses on building automatic speech recognition (ASR) system for Marathi language (M-ASR) using Hidden Markov Model Toolkit (HTK). The M-ASR system gives the detail about experimentation and implementation using the HTK Toolkit. In this work total 106 speaker independent Marathi isolated words were recognized. These unique Marathi words are used to train and evaluate M-ASR system. The speech corpus (database) is created by us using isolated Marathi words uttered with mixed gender people. The system uses Mel Frequency Cepstral Coefficient (MFCC) for the purpose of extracting features using Gaussian mixture model (GMM). Viterbi algorithm based on token passing is used for decoding to recognize unknown utterances. The proposed M-ASR system is speaker independent. The proposed system has reported 96.23% word level recognition accuracy.

Object identification is more significant application in Deep learning innovation which is recognized by areas of strength for its element learning and component portrayal contrasted and the customary article acknowledgement techniques. The paper presents the old-style strategies for object acknowledgement and elucidates the connection and contrast between the traditional techniques and the Deep learning strategies in object discovery. It presents the development of the article acknowledgement techniques in light of profound learning and explains the most ordinary strategies these days in the article location through Deep learning. In this paper we developed a model to check the accuracy based on the Deep Learning techniques to detect the object present within an image and enhance the quality.