Search

Automation in every domain is today’s need, so in question paper setting also. The proposed system is automation of paper setting and identification of difficulty level of questions and question papers, which uses blooms taxonomy categorization to classify the questions in easy, medium and high categories. By using blooms taxonomy, the difficulty level of question can be identified, which is helpful in generating qualitative question paper. While categorizing questions, system uses natural language pre-processing techniques, which are understood by computer and also uses Latent Dirichlet Allocation algorithm. Based on paper setter’s requirement, paper setting is done; also, if user wants to check difficulty level of any question paper then this is also done by proposed system using NLP. The paper setting using difficulty measure is right benchmark for evaluating student’s efficiency. Automation of paper setting improves the speed, and difficulty level module improves the quality.

Background: Presentation of sepsis is dependent on synthesis of varied clinical information, making identification of septic patients challenging. Common practice is to identify sepsis through a manual screening tool, which may miss opportunities for early sepsis detection. Electronic screening efforts requiring additional documentation by providers may not integrate easily into provider workflow. Objectives: To develop an automated sepsis risk screening tool in the electronic health record that would accurately identify patients at risk for sepsis without requiring additional documentation. Methods: Criteria in the manual screening tool were mapped to standard documentation routinely entered in the electronic health record (Epic Systems, Corp.). Data elements were scored electronically at arrival and every 15 minutes during their encounter from the medical history, medication record, vital signs, and physical assessment (Fig. 1). Scores that exceeded a predefined sepsis risk threshold triggered a Best Practice Advisory, which alerted bedside staff to perform sepsis huddles and consider appropriate interventions. Statistical comparison of the automated tool to the manual process was completed by two-tail paired t test. Results: In an 8-week testing period, the automated sepsis risk screening tool identified 100% of patients flagged by the manual process (N = 29) (Table 1). The electronic tool identified sepsis patients, on average, 68 minutes earlier. This was statistically significant (P < 0.001). Conclusions/Implications: The automated sepsis risk screening tool is as accurate as a validated manual process and alerted bedside clinicians earlier. Deployment has potential to improve timely sepsis detection and management of patients without requiring additional documentation by provider.

Automation is at the core of modern industry. It aims to increase production rates, decrease production costs, and reduce human intervention in order to avoid human mistakes and time delays during manufacturing. On the other hand, human assistance is usually required to customize products and reconfigure control systems through a special process interface called Human Machine Interface (HMI). Machine Learning (ML) algorithms can effectively be used to resolve this tradeoff between full automation and human assistance.This paper provides an example of the industrial application of ML algorithms to help human operators save their mental effort and avoid time delays and unintended mistakes for the sake of high production rates. Based on real-time sensor measurements, several ML algorithms have been tried to classify paper rolls according to paper grammage in a white paper mill. The performance evaluation shows that the AdaBoost algorithm is the best ML algorithm for this application with classification accuracy (CA), precision, and recall of 97.1%. The generalization of the proposed approach for achieving cost-effective mills construction will be the subject of our future research., 9 pages

The objective of this research is to design an approach for selecting appropriate solutions to efficiently and productively reduce environmental pollution and water, energy and raw material consumption in a recycled paper manufacturing facility, the paper mill in Latif, Iran. To reach this goal, we develop a decision-making model using the Analytic Hierarchy Process (AHP). The proposed model provides a framework for cleaner production implementation specific to the paper industry. Pairwise comparisons of the results of designed questionnaires resulted in the hierarchy model being divided into five major groups: production modification; process change; preventive maintenance; renewal; and manpower resources.These five groups include 31 sub-criteria and seven alternatives. The results showed that process change yields the highest priority among the main criteria and that industrial automation and its technology level gives the highest priority among sub-criteria. Regarding the alternatives, reductio...

The results of analyzing the process of paper manufacture on the paper machine (PM) in terms of automatic control problems are given in the paper. They are presented in the form of parametric schemes of each part of the PM indicating the main input and output variables, as well as control and perturbation effects. Wire, drying and press parts, sections of pulp preparation, wet washing, screening, inlet and calender are considered in the paper. Construction of parametric schemes is the first step in creating an automatic control system. Based on the obtained results, it is concluded that the most attention from the point of view of control should be given to the PM drying section and calender. The conclusion is associated with a large number of perturbations, acting on these processes and negatively affecting the quality of the finished products but also lead to increasing production costs. With the set parameters of the PM processes, it is possible to formulate an optimal control problem that is the next step in creating an automatic control system.

Water soluble pullulan films were formatted into paper-based microfluidic devices, serving as a controlled time shutoff valve. The utility of the valve was demonstrated by a one-step, fully automatic implementation of a complex pesticide assay requiring timed, sequential exposure of an immobilized enzyme layer to separate liquid streams. Pullulan film dissolution and the capillary wicking of aqueous solutions through the device were measured and modeled providing valve design criteria. The films dissolve mainly by surface erosion, meaning the film thickness mainly controls the shutoff time. This method can also provide time-dependent sequential release of reagents without compromising the simplicity and low cost of paper-based devices.

Nowadays, in this digital word text is one the vital knowledge carrier. Text can be perceived as structure of symbols which is accustomed to convey, transfer, or inherit civilization and way of life. It is among the major influential novelties of man kind; text performs crucial role in existence of human being. Text can be seen in all places in our day-to-day life. Thus, obtaining data from text which is present in scenarios of our day-to-day life is important. Processing the text that is present in natural scenario has two primary tasks namely Text identification and Text acknowledgment. The method of taking out text body from the natural scene is called as Text detection whereas the procedure of categorizing the characters from scene image which form expressive words is called as Text recognition. There are so many applications of scene text detection in different fields, e.g., intelligent inspection, blind navigation, industrial automation, automotive assistance, robot navigation, instant translation, automatic driving and product search, etc. Thus, street view text detection and recognition have allured rising research consideration in past few years. This paper as first step focuses on Text detection methods for natural scene images.

Modern society is deeply intertwined with modern technology, and the reaches of the internet extend into nearly all aspects of life. As more and more of our personal data is stored electronically, the risks of losing that data or having others access that data grows increasingly dangerous. Whether it is financial information or photo albums, internet users have the right to protect their data at all costs, but this may be easier said than done. To dispel attacks, companies and government agencies collect data and employ surveillance measurements to ensure that data is being protected and virtual services are being used properly, but the balance between effective security and over-intrusive surveillance is hard to maintain. With better policy and the introduction of successful cybersecurity programs driven by machine learning, cyberattacks may be better protected against without any infringements upon user privacy. The technical component of this project aims to investigate how machine learning can improve security in electronic systems. By having algorithms or procedures that can easily and accurately identify and prevent cyberattacks, systems can better protect data and prevent intrusions. Using the MITRE CALDERA cybersecurity framework, advisories and defenders can be simulated to model cyber attacks and prepare defenses for these attacks. Adversaries are equipped with abilities that can be set up with tactics, techniques, and procedures, or TTPs, detailing how an attack will be carried out. Agents, which can be advisories or defenders, can be placed in groups with each group working to attack or defend. The defending agents are equipped with abilities to defend against various attacks and the abilities are constantly updated to better recognize and prevent attacks. The STS component of this project focuses on the use of surveillance in remote work environments. As many companies take steps to protect their data, employees are increasingly being monitored to ensure they are not putting company data at risk and to encourage maximum productivity. This has raised concerns as employees feel as though their privacy is being violated and they must compete with employers to find a balance. Due to the COVID-19 pandemic, many companies shifted to remote work environments, and with a lack of in-person monitoring, many employers introduced virtual surveillance measures that extended the reach of employers further into the personal lives of employees. Policy changes both on the company level and government level must be implemented to protect the rights of employers and clearly define the jurisdictions of private and public corporations. This project aims to highlight the flaws and successes of current cybersecurity protocols a well as investigate the ethical issues related to protective surveillance measures. These issues must be considered not only by engineers in the related fields but also by policy makers and employers so that privacy on the internet can be strictly maintained. I would like to extend my gratitude to my advisors, Professors Peter Norton, Sean Ferguson, and Jack Davidson for their support and guidance throughout this project.

Paper-based digital microfluidic biochips (PDMFBs) are becoming highly effective among the microfluidic platforms due to its low-cost and in-place fabrication. The designed electrodes and wiring can be fabricated on a piece of paper by an inkjet printer and conductive ink containing carbon nanotube particles (CNTs). However, due to induced control interference, the wires cannot pass by an arbitrary electrode. Each wire that is to be routed possesses its conflict electrode group, which must be avoided for a feasible droplet movement. This paper presents a fluid-control co-design considering several important cost-driving issues like minimization of schedule length, control pin count, and wirelength, together with congestion-free and conflict-free wiring. Observably, design gaps exist among the sub-tasks of the fluid-level, control-level, and fluid-control as a whole, due to their separate considerations. It indeed introduces many design cycles lengthening the design process, and thus increases the overall cost. In this context, this work integrates the sub-tasks as a prior need to obtain a low cost and efficient platform. Several benchmarks have been studied to evaluate the performance.

The Paper-to-Reviewer Assignment Problem is a prevalent and challenging problem within the scientific community and scientific conferences. The large number of reviewers and submitted papers, as well as constraints on reviewers’ loads and papers’ coverage needs, make difficult and very tedious for the committee chair to manually assign submitted papers to suitable reviewers. Any automation of this task must tackle three main issues: modeling reviewers’ and papers’ profiles, estimating the relevance between a paper and a reviewer and finally finding an appropriate assignment.In this paper, we are interested in the two first issues. We propose to build the reviewers profiles on the basis of both pedagogical and research documents, then a formula that takes into account both topics similarity and common references is proposed. Experimental results compared to ground-truth values extracted from ICA2IT’19 data show a good performance of this approach.

Despite tremendous advancement of digital IC design automation tools over the last few decades, analog IC layout is still heavily manual which is very tedious and error-prone. This paper will first review the history, challenges, and current status of analog IC layout automation. Then, we will present MAGICAL, a human-intelligence inspired, fully-automated analog IC layout system currently being developed under the DARPA IDEA program. It starts from an unannotated netlist, performs automatic layout constraint extraction and device generation, then performs placement and post-placement optimization, followed by routing to obtain the final GDSII layout. Various analytical, heuristic, and machine learning algorithms will be discussed. MAGICAL has obtained promising preliminary results. We will conclude the paper with further discussions on challenges and future directions for fully-automated analog IC layout.

Registered as Defensive Paper of the Intellectual Property Invention "Clustering Techniques for Revenue Accounting Error-Handling Automation", Patent number ID2326WW00, Intellectual Property Board, Amadeus S.A.S., Sophia Antipolis, France. July, 2020.; International audience; We present a semi-supervised ensemble clustering framework for identifying relevant multi-level clusters, regarding application objectives, in large datasets and mapping them to application classes for predicting the class of new instances. This framework extends the MultiCons closed sets based multiple consensus clustering approach but can easily be adapted to other ensemble clustering approaches. It was developed to optimize theAmadeus S.A.S revenue accounting workflow management. Revenue accounting in travel industry is a complex task when travels include several transportations, with associated services, performed by distinct operators and on geographical areas with different taxes and currencies for example. Preliminary results show the relevance of the proposed approach for the automation of workflow anomaly corrections.

In education field, the importance of education system is the part of the most significant aspect of our social orders at all levels. Thus, a student's performance can only be tested by providing them with an examination or assessment. Examinations are very crucial as it is an activity which is performed to evaluate the performances of students. Besides students performances, the quality of the examination questions prepared would also determine the quality of the students but the most important yet challenging task is the creation of the examination questions paper which is known to be very tedious, time consuming and expensive therefore it should be prepared with full focus, right formatting and with the right selection of questions.

Data confidentiality, integrity and authentication are security properties which are often enforced with the generation of secure channels, such as Virtual Private Networks, over unreliable network infrastructures. Traditionally, the configuration of the systems responsible of encryption operations is performed manually. However, the advent of software-based paradigms, such as Software-Defined Networking and Network Functions Virtualization, has introduced new arms races. In particular, even though network management has become more flexible, the increased complexity of virtual networks is making manual operations unfeasible and leading to errors which open the path to a large number of cyber attacks. A possible solution consists in reaching a trade-off between flexibility and complexity, by automatizing the configuration of the channel protection systems through policy refinement. In view of these considerations, this paper proposes a preliminary study for an innovative methodology to automatically allocate and configure channel protection systems in virtualized networks. The proposed approach would be based on the formulation of a MaxSMT problem and it would be the first to combine automation, formal verification and optimality in a single technique.

Fog-based radio access network (F-RAN) have the potential to meet the stringent latency and bandwidth requirements of next-generation mobile networks. However, the current F-RAN is implemented mainly by utilizing dedicated hardware and does not leverage on the available large number of distributed edge devices. This motivated us to recently propose the idea of an opportunistic fog RAN (OF-RAN) which introduces the concept of a virtual fog access points (v-FAP) formed by two or more local edge devices (referred as service nodes), and monitored by physical FAPs (referred to as seed nodes). Some of the challenges arising from having many devices opportunistically deployed to serve client nodes are accessing, forming and managing them. In this paper, we propose to harness blockchain technology for assigning roles and permissions to service nodes and seed nodes in our OF-RAN, and supporting the formation and management of v-FAPs. For this blockchain architecture, the distributed formation and management algorithm (DFMA) is implemented as a script in the smart contract. The blockchain then executes the logic of the smart contract while considering the resources from all service nodes participating in the system. By integrating OF-RAN with blockchain systems and smart contracts, it is possible to provide verifiable automation of physical processes involving different nodes.

The design and engineering of solutions based on sensor data is currently based on legacy low-level sensor communication. The transition to IoT-based sensors has started to provide IP-based sensor communication based on service concepts. This requires radically different solution architectures and the associated engineering process. Here, IoT sensors will provide services instead of a communication interface, as a means of integrating into application solutions. It is further anticipated that the size of the solution will reach well beyond that of the current legacy automation implementations.This paper addresses architectural concepts for very large-scale SoS- and IoT-based solutions. Architecture concepts and their associated engineering concepts are considered. IoT and SoS evolvability, run-time dynamics, scalability and segmentation, run-time engineering, self-engineering, self-mitigation, and machine-to-machine business are some examples of such concepts. These concepts must be coupled with the requirements such as the management of IoT/SoS solution functionality, security, safety, maintenance, and evolution.

Logical frameworks like LF allow to specify the syntax and (natural deduction) inference rules for syntax/proof-checking a wide variety of logical systems. A crucial feature that is missing for prototyping logics is a way to specify basic proof automation. We try to alleviate this problem by generating \(\lambda \)Prolog (ELPI) inference predicates from logic specifications and controlling them by logic-independent helper predicates that encapsulate the prover characteristics. We show the feasibility of the approach with three experiments: We directly automate ND calculi, we generate tableau theorem provers and model generators.

This paper presents enhancements of Service Oriented SystemJ (SOSJ) framework, which extends a system-level language based on GALS model of computation SystemJ with services, into a new programming paradigm amenable for designing dynamic distributed automation systems such as reconfigurable manufacturing systems. The new paradigm combines correct-by-construction software systems development available in SystemJ with the dynamic features of service oriented architecture. The new approach introduces macro states into fundamental concurrent and distributed entities of SOSJ, called clock domains, which address typical behaviors in dynamic distributed systems. We showcase the use of the new paradigm on an example of reconfigurable manufacturing scenarios in a dynamic manufacturing system.