No abstract

This research involves the usage of Machine Learning technology and Natural Language Processing (NLP) along with the Natural Language Tool-Kit (NLTK). This helps develop a logical Text Summarization tool, which uses the Extractive approach to generate an accurate and a fluent summary. The aim of this tool is to efficiently extract a concise and a coherent version, having only the main needed outline points from the long text or the input document avoiding any type of repetitions of the same text or information that has already been mentioned earlier in the text. The text to be summarized can be inherited from the web using the process of web scraping or entering the textual data manually on the platform i.e., the tool. The summarization process can be quite beneficial for the users as these long texts, needs to be shortened to help them to refer to the input quickly and understand points that might be out of their scope to understand.

No abstract

Software is a generic term for organized collections of computer data and instructions; it is amalgamation of machine understandable instructions that preside over the processor of computer to perform itemized operations. The importance of software accompanies the effect of its malleability towards software development life cycle. Antecedent models such as Prototyping model, Incremental model, waterfall model, Transformation model, RAD model, Spiral model Model, Agile models and so on were found to be more efficient during its inception phase which later leaded to have some barricades such as risk factor, cost factor, time complexity and so on. The present research emphasizes a new model i.e. Software Augmentation Isochronism model in order to perturb those barricades present with the previous approaches. This paper clenches a prominent model with features such as classified exigency, bilateral databases, mini marketing, testimonial access and significant testing methodologies in order to produce efficient software which decreases the risk factors and decreases the costs to develop the software and thus finally provides a complete complacency and thesis of assurance between customer and company.

This paper introduces a new decision tree algorithm Diabetes Prediction Algorithm (DPA), for the early prediction of diabetes based on the datasets. The datasets are collected by using Internet of Things (IOT) Diabetes Sensors, comprises of 15000 records, out of which 11250 records are used for training purpose and 3750 are used for testing purpose. The proposed algorithm DPA yielded an accuracy of 90.02 %, specificity of 92.60 %, and precision of 89.17% and error rate of 9.98%. further, the proposed algorithm is compared with existing approaches. Currently there are numerous algorithms available which are not complete accurate and DPA helps.