Search

Automatic question-answer pair generation is gaining importance because of the time being saved when compared to manual creation of questions. It not only saves time but also reduces the effort involved in the manual question generation process. It is useful in many fields such as school assignments, law practicing, self-assessments and many more. Our main objective is to create wh-type questions from a paragraph and find its accurate answer as well. This research work introduces a Question Answer Generation (QAG) system by using a deep learning approach for combining Answer Extraction (AE), Question Generation (QG) and Question Answering (QA) models. This research work aims to use a pre-trained language model - Bidirectional Encoder Representation from Transformers (BERT) and fine-tune it as per the proposed research objective.

In this paper, an automatic system is proposed to generate different kinds of questions and answers from the input text. Question answer generation systems have been an interesting field of research for over decades. From generating questions for educational purposes to preparing answers to questions that could be asked in a legal proceeding, the purpose of question answer generation(QAG) systems is to reduce the tedious task of going through large texts. In our system, question-answer pairs from a given input text are generated using linguistic and statistical knowledge of text. Initially, those sentences are identified from the input text on which questions can be framed and in further steps, identified sentences are ranked in an order of importance. Built specifically for assessment purpose, the system generates multiple choices, fill-ups, true-false, binary and wh type questions based on high ranked sentences in the last step. In performance evaluation of the proposed system, it is found that the system is performing well and it is close to state of art research in the standard linguistic approach.

The objective of this article is to develop an intrusion detection model aimed at distinguishing attacks in the network. The aim of building IDS relies on upon preprocessing of intrusion data, choosing most relevant features and in the plan of an efficient learning algorithm that properly groups the normal and malicious examples. In this experiment, the detection model uses an ensemble approach of supervised (SVM) and unsupervised (K-Means) to detect the patterns. This technique first divides the data and forms two clusters as per K-Means and labels the clusters using the Support Vector Machine (SVM). The parameters of K-Means and SVM are tuned and optimized using an intrusion dataset. The SVM provides up to 88%, and K-Means provides up to 83% accuracy individually. However, the ensemble of K-Means and SVM provides more than 99% on three benchmarked datasets in less time. The SVM only classifies three instances of each cluster randomly and labels them as per a majority voting approach. The proposed approach outperforms compared to earlier ensemble approaches on intrusion datasets.