In today's era, women still feel scared to step out of their homes at late nights due to the fear of sexual harassment. This problem is compounded when one shifts the focus from urban to rural sector and interior areas of villages where there is erratic electric supply and poor cellular network connectivity. There are many smartphone based solutions, but the availability of smartphone and cellular network in rural areas is unreliable. This paper proposes a system, by creating a wireless network using IoT technology with a portable device for alerting the concerned authorities to prevent any mishap. Women will be provided with a beacon device consisting of a help button. In case of any emergency, the beacon information will reach the central stations and an alarm will be triggered at the prominent places of the village. One can find out the location of the victim based on her proximity to the nearest access point.

In recent years, biometrics using facial recognition has become an important part of many industries like security, retail, marketing, health-care, etc. Internet of Things has played a pivotal role in enhancing and automating technology for practical applications. However, there exist some issues with implementing these systems practically. In this paper, an attendance management system is proposed which can detect and recognize faces of an entire class from a few pictures and mark the attendance of the recognized students. IoT cameras have been used instead of a smartphone camera to increase coverage and reduce the number of missed attendances. In addition to reducing manual work, it has also been ensured that this system is completely secure. The system proposed mainly comprises of IoT Camera Module, custom back-end server and separate Android apps for teachers and students. The system is also able to provide high accuracy at low cost as compared to other other similar applications and hence is suitable for real-life use.

The growth of semantic web has given rise to proliferation of data sources wherein the task of recognizing real world entities and identifying multiple references of the same real world entity becomes an essential task in order to facilitate sharing and integration of data. Due to the heterogeneous nature of data on the semantic web, entities belonging to different sources are compared by assessing the similarity of features that are common in order to identify matches. With the increasing size of data sets Candidate generation methods are generally employed to avoid quadratic time complexity that would otherwise be incurred if pairwise similarity of all entities are computed. Here we propose a novel index based approach for candidate generation and reduction. The evaluation shows that the proposed method scales well and improves recall significantly.

Context-free grammars (CFGs) are a widely used class of formal grammars specially for the parsing of programming languages, files and data streams. Comparison of the structural information represented by different CFG will allow several additional applications in these contexts. Most current CFG comparison techniques focus on the determination of equivalence or non-equivalence rather than on determining the degree of similarity and so exploit only some of the capabilities of CFG. We present an algorithm for CFG Affinity Analysis using automatically and randomly generated test sentences. Our method provides an efficient and simple solution to a problem that has not yet received much attention. We intend to use this approach to develop a system for retrieval of structured data streams like music which can be effectively and sufficiently represented using CFG.

In 2013, authors hypothesized that students would improve if motivated and mentored appropriately to pursue career in the chosen path. The primary goal is to facilitate students' successful transfer in the chosen path by maintaining healthy relevant learning environment. This study evaluates whether the implemented career guidance program in the electronics engineering department increases self-esteem, confidence and skill to pursue career in the chosen path by the student. Results indicate that students show higher levels of academic performance and achievements. In this paper, authors also propose curriculum scheme with choice based career guidance.

Diabetic retinopathy is when damage occurs to the retina due to diabetes, which affects up to 80 percent of all patients who have had diabetes for 10 years or more. The expertise and equipment required are often lacking in areas where diabetic retinopathy detection is most needed. Most of the work in the field of diabetic retinopathy has been based on disease detection or manual extraction of features, but this paper aims at automatic diagnosis of the disease into its different stages using deep learning. This paper presents the design and implementation of GPU accelerated deep convolutional neural networks to automatically diagnose and thereby classify high-resolution retinal images into 5 stages of the disease based on severity. The single model accuracy of the convolutional neural networks presented in this paper is 0.386 on a quadratic weighted kappa metric and ensembling of three such similar models resulted in a score of 0.3996.

E-learning, in 21st century, is considered as the important part for distance learning education. Many content development models are currently available which creates content as per their own standards, but sharing of such content is merely impossible among different learning systems, so, each model should follow some common standards which can be helpful to share content. Authoring tool adhering to international standards can be used to create learning object which can be used by E-learning systems and also help in sharing learning objects. This tool can create content as per required parameter supported by repository which can also used in dynamic linking of different learning objects. This paper describes the need of one such authoring tool which works on principle based on content development model and relation between contents based on keywords and user response.

People for various reasons need to learn new topics and acquire new skills, but they don't want to attend class or to take full e-learning courses. They search the internet and variety of other sources, to swiftly learn what they need to know. For learners, it is difficult to find and arrange this information for effective learning. Knowledge acquired is often detached, forgotten or not well incorporated into practice. These observations are not unique to the working professionals. Learners in colleges and universities also spend considerable amount of effort in search for information. In all cases, content on the web is often not well organized according to learners learning needs thus learners spend considerable amount of time in unproductive interactions. A more flexible approach to satisfy each learner's learning goal and context is required. This research focuses on dynamic sequencing learning object based on learning goal and learning object prerequisite. A dynamic course curriculum is created for each learner with option to choose learning path suitable for the learner.

In this paper, a new indexing technique is proposed which retrieves the proximity of the keyword when exact match is not found. It is based on a signature graph along with n-ary tree which retrieves the relevant information when exact match does not exist. The index structure arranges signatures in aggregation hierarchy which leads to improvement of query evaluation. It also uses an n-ary tree data structure which is a hierarchy of keywords which is used to map in case of inexact match. As a proof of concept, a small application is built for a project manager to organize and retrieve resources efficiently. The application retrieves employee with required skill set and if such person is not available then it locates the closest match for the requirement. It also provides adequate resources to improve skill sets of the closest match selected. Thus, this system provides one point solution for the project manager.

Exhaustive product evolution and testing is required to ensure the quality of product. Regression testing is crucial to ensure software excellence. Regression test cases are applied to assure that new or adapted features do not relapse the existing features. As innovative features are included, new test cases are generated to assess the new functionality, and then included in the existing pool of test cases, thus escalating the cost and the time required in performing regression test and this unswervingly impacts the release, laid plan and the quality of the product. Hence there is a need to select minimal test cases that will test all the functionalities of the engineered product and it must rigorously test the functionalities that have high risk exposure. Test Suite-Test Case Refinement Technique will reduce regression test case pool size, reduce regression testing time, cost & effort and also ensure the quality of the engineered product. This technique is a regression test case optimization technique that is a hybrid of Test Case Minimization based on specifications and Test Case Prioritization based on risk exposure. This approach will facilitate achievement of quality product with decreased regression testing time and cost yet uncover same amount of errors as the original test cases.