1,720,962 research outputs found
Energy-Efficient Cache Localization in Device-to-Device Network
No full textLast few years has seen one of the most important challenges for smart devices that is resource allocation. This has been a major bottleneck as there have been advancements in the cellular technology, especially 5G. With the increase in the enormous amount of number of users, there has been an increase in the demand of services and content within fraction of seconds. This has increased the burden on the network with respect to QoS and QoE to the end users and service providers. Caching the most popular contents on the user’s equipments (UE’s) can solve the problems because contents will be nearer to the users and can be shared using Device-to-Device (D2D) communication without the involvement of the core network. Motivated by this fact, in this thesis we present an extensive survey on the present caching techniques for D2D communication in 5G. Further, we present a model to address the problem of cache location decision. First, we find form a network of users using Matlab. Then we create dataset for predicting the cache locations in a network. Further, we predict the location where the user caches the most accessed content using machine learning classification models. The classification models used are decision tree and random forest. We compare the results obtained from decision tree and random forest classification model. On comparison we observe that the random forest model considering the trust factor between the users has higher accuracy. Then we use iFogSim simulator, to simulate the sending and receiving of contents in the network in a Software Defined Network environment. The metrics used for evaluating the simulations are access delay and energy consumption of the user’s equipment (UE’s). On analyzing the obtained result from simulation, we observe that the access delay is maximum at the users end when the sharing is with the gateway and energy consumption of the UE’s is also the maximum when the contents are accessed from the gateway.TIE
Sentiment Analysis of Twitter Data Using Machine Learning Techniques
No full textMaster of Engineering-Software EngineeringOnline Microblogging on social networks have been used for indicating opinions about
certain entity in very short messages. Existing some popular microblogs like twitter,
facebook etc,in which twitter attains maximum amount of attention in the field of
research areas related to product, movie reviews, stock exchange etc. The research on
sentiment analysis has been going for a long time. Sentiment analysis in present days
becomes the major issue in field of research and technology. Due to day by day increase
in the number of users on the social networking websites, huge amount of data produces
in the form of text, audio, video and images. There is need to do sentiment analysis as
texts in form of messages or posts to find the whether the sentiment is negative, positive
or neutral. We had extracted data from twitter i.e. movie reviews for sentiment prediction
using machine-learning algorithms. We applied supervised machine-learning algorithms
like support vector machines (SVM), maximum entropy and Naïve Bayes to classify data
using unigram, bigram and hybrid i.e. unigram + bigram features. Result shows that SVM
surpassed other classifiers with remarkable accuracy of 84% for movie reviews
Efficient Rule-Based Grammar Checker with Word Sequencing
No full textMaster of Engineering -CSEA grammar checker is a program which is used to correct the grammar of a sentence
or a paragraph. These days, informal writing (communication on social websites and
chat applications) has gained a lot of popularity. In informal writing there is no
requirement of grammar to be correct. However when we come to formal writing
such as research papers, emails, patents etc. then it is required that, English written in
these formal writings should be free of grammar errors. Moreover, error free text
explains the idea or thought more clearly.
In this thesis we present a rule-based grammar checker, which is capable of correcting
the word order of a sentence in English language. Our tool first identifies the tense of
the input sentence. For every tense we have different categories of sentences. Then
based on the tense, it further categorizes the sentence into the category it belongs, and
depending on the category of the sentence, corresponding function is called which
corrects the sentence. The correction is done with the help of certain rules that we
have defined for every category of sentence. The rules are based on the word order of
the sentence. Our tool also incorporates the work of existing grammar checkers. This
work further checks the grammar of the sentence. If the grammar of sentence is
incorrect, then it is corrected after correcting the word order of the sentence.
For evaluating the performance of our tool, we have compared it with already existing
nine grammar checkers. Results show that the accuracy of our grammar checker is
better than these grammar checkers. Moreover, our tool is able to correct the incorrect
order of the sentence which in not done by the previous researchers
Cognitive Analysis Based System for Effective Online Learning
No full textOnline learning has brought about a transformative shift in the field of education worldwide. It has revolutionised traditional educational practices, offering new possibilities and opportunities for learners, educators, and institutions. Online learning has helped bridge the gap in access to education. It has reached individuals in remote areas or underserved communities who previously faced limited educational opportunities. By removing geographical barriers, online learning has made education more inclusive, ensuring that learners from all backgrounds can access quality education. Students from all corners of the world can now enrol in courses and programs offered by prestigious universities and institutions, democratising education on a global scale. Online learning has emerged as a crucial tool during the COVID-19 pandemic, allowing individuals to continue their education remotely. With the closure of physical classrooms and educational institutions, online learning has stepped in to ensure continuity in learning. It has allowed students to stay connected with their educational pursuits, despite unprecedented challenges. By using technology and digital platforms, online learning has enabled students to access educational resources, interact with teachers and peers, and participate in virtual classes. Despite the limitations imposed by the pandemic, it has provided an adaptable solution that has allowed the education system to continue functioning. Virtual classes and online courses have become the primary mode of education worldwide, with platforms like Coursera and Udemy offering free course enrolments during the pandemic.
Despite the widespread adoption of online learning, it still needs to improve to ensure learners remain attentive throughout online learning sessions like in physical classroom settings. Monitoring student attentiveness accurately in online learning is difficult for teachers, as they cannot rely on direct observations about what happens in the physical classroom environment. In traditional classes, teachers can identify distracted students through observation, differentiating those who are actively engaged from those who are disengaged. But these observations are only possible in the traditional classroom environments or, in other words, face-to-face interaction scenarios are not available in the online learning environment. Students generally fail to maintain their learning level without immediate feedback or support. As a consequence of it, learners prefer to leave the online course or online class mid-way. That is why the major issue educators and policymakers face in online learning medium is high drop-out rates. Educators are striving to find ways to keep learners engaged and motivated during online learning sessions, highlighting the need to design engaging
and student-centric online learning environments. To address these challenges and enhance the effectiveness of online learning, it is important to monitor, detect, and predict the cognitive state of online learners. The cognitive state refers to the mental and psychological processes that influence a learner’s engagement, attention, and overall learning performance during the learning process. By understanding the cognitive state of learners, educators can adapt their teaching strategies, provide timely feedback, and other personalised support to enhance the learning experience.
Various methods are employed for this purpose. Self-reporting is one approach where learners provide feedback on their understanding and attentiveness after online course completion. However, this method can be subjective and unreliable as it is human-biased and lacks the facility to provide immediate support to the learners if they are unsatisfied. Another approach involves the analysis of learning analytics data. By tracking learner’s digital footprints, such as their interactions with online platforms, task completion rates, and time spent on activities, valuable insights can be gained. But again, this does not fully reflect the learner’s attentiveness and does not tell the instance where online learners face issues. That’s why there is a growing need to develop effective systems that can monitor the cognitive state of online learners in real-time. Therefore, this thesis presents a cognitive analysis-based system specifically designed for online learning environments.
The proposed system aims to enhance learner engagement, attention, and overall learning outcomes by utilising various techniques to analyse and interpret their cognitive states. In this research, two techniques, namely facial cues and EEG signals, are employed to monitor and analyse the cognitive behaviour of learners during online learning sessions. Facial cues include facial emotions, eye tracking, and head movements, which are captured and processed in real-time to assess the learner’s cognitive states. Monitoring facial cues provides a direct means to recognise the learner’s focus during online learning, similar to the observation in a physical classroom environment. The face, eye, and head movements serve as powerful visual indicators to interpret an individual’s intentions while learner’s engaged in an online learning session. The changes in facial emotions and eye and head movements can be considered as the learner’s responses to the delivered content in the online learning environment. To capture these cues, a built-in web camera is used to obtain real-time details of the face, eye, and head. Traditional Deep CNN models for facial emotion recognition are employed, and a CNN-based model is also proposed. The publicly available benchmarked datasets are used for training these models. Additionally, an in-house collected face dataset is utilised for testing purposes. The detection of eye-blinking patterns and head movements is achieved using a landmark approach. The outcomes
of these face cues are combined to predict the final cognitive state of the online learner, whether they are attentive or distracted. The performance of the system is evaluated by comparing the standard baseline deep CNN models with the proposed model for facial emotion recognition. The experimental analysis demonstrates that the proposed model outperforms traditional deep learning algorithms for facial emotion classification. Furthermore, a comparative analysis of the proposed cognitive analysis system based on facial cues is conducted to evaluate its effectiveness against state-of-the-art models.
This thesis also considers EEG signals for cognitive state prediction. EEG signals are utilised to measure learner’s brainwave activities, providing valuable insights into their cognitive workload and attention levels. By using EEG-based technology, instructors/teachers can observe a learner’s cognitive load, which refers to the level of mental effort exerted during learning tasks, without disrupting the online learning process. A Bluetooth-enabled single-electrode EEG device is employed to capture the electrical signals generated by the flow of electrons across neurons during learning tasks. EEG is considered the most effective sensor for measuring cognitive load and attentiveness in the online learning environment. It identifies fluctuations in signal levels, processes the data, and reflects different levels of alertness. EEG provides unbiased information about learner’s cognitive states, making it suitable for real-time analysis. The EEG device measures the electrical activity of neurons in the brain cortex using specific electrodes and categorises the activity into different frequency bands. The brainwave signals obtained from EEG recordings are categorised into five frequency bands: Delta, Theta, Alpha, Beta, and Gamma, each associated with specific mental activities or states. Machine learning algorithms are employed to analyse the collected EEG signal data and predict the learner’s cognitive states. These algorithms are trained using a combination of publicly available datasets and the in house EEG data collected in this research. Performance metrics such as accuracy, precision, recall, and F1 score are utilised to evaluate the effectiveness of the proposed system in predicting and classifying cognitive states. Furthermore, a comparative analysis of the proposed cognitive analysis system based on EEG signals is conducted to assess its effectiveness compared to state-of-the-art models. The implemented cognitive evaluation for face cues and EGG signals are represented separately with the help of a web application. The web application for the proposed smart education system is also presented in this thesis. The Web Application comprises a range of features that are thoroughly discussed and incorporated into the system. These features are designed to enhance the overall functionality and effectiveness of the smart education system, providing a comprehensive and user-friendly platform for online learning
An Optimized Approach for Energy Consumption of Smart Devices in Fog Computing using Computational Intelligence Techniques
No full textPhD ThesisTo address the frequent overloading of fog nodes due to the increasing demand for IoT applications, an ensemble approach was employed to classify host load status into underloaded, balanced and overloaded categories. This work introduces an innovative reliability framework that encompasses multiple implementation phases. The process begins with the generation of virtual machines via the command line using various random settings.
Various parameters such as CPU utilization, number of CPU cores, RAM, memory allocation, memory availability, disk I/O, and network I/O were analyzed to better understand host workload. Three case studies with varying numbers of virtual machines (VMs) were conducted on two different platforms for load prediction.
A total of ten machine learning models were employed to construct an ensemble model, which ultimately yielded optimal and accurate results for classifying host load. All models are evaluated for their precision, recall, and accuracy. Various pre-processing techniques such as normalization, transformation, principal component analysis (PCA), outlier removal are applied on the generated dataset and various models are compared.
It was observed that applying normalization to a dataset improved the performance of the models. Four models—Random Forest (RF), AdaBoost (AB), Gradient Boost (GB), and Decision Tree (DT)— performed equally well across all three case studies with normalized datasets. However, our proposed ensemble model performed marginally better than these individual models and it achieved nearly 82% accuracy in correctly
classifying host load.
As the next major revolution in cloud and fog computing environment, container migration and containerization have emerged as key advancement. Fog computing and mobile edge cloud necessitate the transfer of containers from overloaded hosts to new hosts to ensure adequate resources for executing consumer applications at the network edges. Despite the growing popularity of containers, algorithms to manage the excessive energy consumption of hosts have not been thoroughly investigated. Moreover, optimizing the energy consumption efficiency of hosts remains a critical and challenging task.
In order to address the critical issue of reducing energy consumption in fog computing environment, the study moved beyond traditional virtualization techniques, which have a high computational overhead and are less suitable for fog devices. Containers, known for their efficiency in encapsulating fog services, were used instead. A container selection algorithm was introduced to identify containers for migration when a host becomes overloaded.
Further, an energy-efficient container migration strategy was implemented using a dynamic inertia weight-based particle swarm optimization (DIWPSO) algorithm. This strategy aimed to balance the load and reduce energy consumption by migrating containers from overloaded hosts. Experimental results demonstrated that the DIWPSO algorithm significantly reduced energy consumption by 10.89% and achieved load balancing at a lower migration cost compared to traditional meta-heuristic solutions such as PSO, ABC, and E-ABC.
Additionally, The study developed a multivariate time series ensemble model for load prediction on hosts, utilizing anomaly detection techniques to forecast CPU utilization in the near future. Based on these predictions, resource utilization for container management was forecasted, determining the number of hosts needed to support the load of running containers.
Anomaly detection techniques were employed to reduce redundancy in generated data and address inconsistencies in load prediction due to the large volume of data. A predictive model with variable load patterns can better estimate future resource needs, which is crucial for capacity planning, meeting service-level goals, and achieving energy efficiency.
Various time series-based models were used for workload prediction, and the top three models were selected based on their TOPSIS scores to develop the ensemble model. To ensure the efficiency of the proposed model, Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), and accuracy were evaluated and compared with other existing stateof-the-art models. The results demonstrated that the proposed ensemble model exhibited higher accuracy in workload prediction compared to current state-of-the-art models, achieving the lowest Mean Absolute Percentage Error (MAPE) and providing an accuracy of approximately 88%.
In conclusion, by integrating advanced machine-learning models for load prediction with an optimized container migration strategy, the study effectively enhanced resource utilization and energy efficiency in fog computing environment. This comprehensive approach successfully addressed the dual challenges of load balancing and energy consumption, providing a robust solution for managing the increasing demands of IoT applications
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Get PDF“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Get PDFWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Get PDFWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
