Cyberbullying Detection based on Semantic-Enhanced Marginalized Denoising Auto-Encoder

Cyberbullying Detection based on Semantic-Enhanced Marginalized Denoising Auto-Encoder

Cyberbullying Detection based on Semantic-Enhanced Marginalized Denoising Auto-Encoder

Cyberbullying Detection based on Semantic-Enhanced Marginalized Denoising Auto-Encoder

ABSTRACT:

As a side effect of increasingly popular social media, cyberbullying has emerged as a serious problem afflicting children, adolescents and young adults. Machine learning techniques make automatic detection of bullying messages in social media possible, and this could help to construct a healthy and safe social media environment. In this meaningful research area, one critical issue is robust and discriminative numerical representation learning of text messages. In this paper, we propose a new representation learning method to tackle this problem. Our method named Semantic-Enhanced Marginalized Denoising Auto-Encoder (smSDA) is developed via semantic extension of the popular deep learning model stacked denoising autoencoder. The semantic extension consists of semantic dropout noise and sparsity constraints, where the semantic dropout noise is designed based on domain knowledge and the word embedding technique. Our proposed method is able to exploit the hidden feature structure of bullying information and learn a robust and discriminative representation of text. Comprehensive experiments on two public cyberbullying corpora (Twitter and MySpace) are conducted, and the results show that our proposed approaches outperform other baseline text representation learning methods.

EXISTING SYSTEM:

  • Previous works on computational studies of bullying have shown that natural language processing and machine learning are powerful tools to study bullying.
  • Cyberbullying detection can be formulated as a supervised learning problem. A classifier is first trained on a cyberbullying corpus labeled by humans, and the learned classifier is then used to recognize a bullying message.
  • Yin et.al proposed to combine BoW features, sentiment features and contextual features to train a support vector machine for online harassment detection.
  • Dinakar et.al utilized label specific features to extend the general features, where the label specific features are learned by Linear Discriminative Analysis. In addition, common sense knowledge was also applied.
  • Nahar et.al presented a weighted TF-IDF scheme via scaling bullying-like features by a factor of two. Besides content-based information, Maral et.al proposed to apply users’ information, such as gender and history messages, and context information as extra features

DISADVANTAGES OF EXISTING SYSTEM:

  • The first and also critical step is the numerical representation learning for text messages.
  • Secondly, cyberbullying is hard to describe and judge from a third view due to its intrinsic ambiguities.
  • Thirdly, due to protection of Internet users and privacy issues, only a small portion of messages are left on the Internet, and most bullying posts are deleted.

PROPOSED SYSTEM:

  • Three kinds of information including text, user demography, and social network features are often used in cyberbullying detection. Since the text content is the most reliable, our work here focuses on text-based cyberbullying detection.
  • In this paper, we investigate one deep learning method named stacked denoising autoencoder (SDA). SDA stacks several denoising autoencoders and concatenates the output of each layer as the learned representation. Each denoising autoencoder in SDA is trained to recover the input data from a corrupted version of it. The input is corrupted by randomly setting some of the input to zero, which is called dropout noise. This denoising process helps the autoencoders to learn robust representation.
  • In addition, each autoencoder layer is intended to learn an increasingly abstract representation of the input.
  • In this paper, we develop a new text representation model based on a variant of SDA: marginalized stacked denoising autoencoders (mSDA), which adopts linear instead of nonlinear projection to accelerate training and marginalizes infinite noise distribution in order to learn more robust representations.
  • We utilize semantic information to expand mSDA and develop Semantic-enhanced Marginalized Stacked Denoising Autoencoders (smSDA). The semantic information consists of bullying words. An automatic extraction of bullying words based on word embeddings is proposed so that the involved human labor can be reduced. During training of smSDA, we attempt to reconstruct bullying features from other normal words by discovering the latent structure, i.e. correlation, between bullying and normal words. The intuition behind this idea is that some bullying messages do not contain bullying words. The correlation information discovered by smSDA helps to reconstruct bullying features from normal words, and this in turn facilitates detection of bullying messages without containing bullying words.

ADVANTAGES OF PROPOSED SYSTEM:

  • Our proposed Semantic-enhanced Marginalized Stacked Denoising Autoencoder is able to learn robust features from BoW representation in an efficient and effective way. These robust features are learned by reconstructing original input from corrupted (i.e., missing) ones. The new feature space can improve the performance of cyberbullying detection even with a small labeled training corpus.
  • Semantic information is incorporated into the reconstruction process via the designing of semantic dropout noises and imposing sparsity constraints on mapping matrix. In our framework, high-quality semantic information, i.e., bullying words, can be extracted automatically through word embeddings.
  • Finally, these specialized modifications make the new feature space more discriminative and this in turn facilitates bullying detection.
  • Comprehensive experiments on real-data sets have verified the performance of our proposed model.

SYSTEM ARCHITECTURE:

Cyberbullying Detection based on Semantic-Enhanced Marginalized Denoising Auto-Encoder

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

 

  • System : Pentium Dual Core.
  • Hard Disk : 120 GB.
  • Monitor : 15’’ LED
  • Input Devices : Keyboard, Mouse
  • Ram :

SOFTWARE REQUIREMENTS:

 

  • Operating system : Windows 7.
  • Coding Language : JAVA/J2EE
  • Tool : Netbeans 7.2.1
  • Database : MYSQL

REFERENCE:

Rui Zhao and Kezhi Mao, “Cyberbullying Detection based on Semantic-Enhanced Marginalized Denoising Auto-Encoder”, IEEE Transactions on Affective Computing, 2016.

Energy and Memory Efficient Clone Detection in Wireless Sensor Networks

Energy and Memory Efficient Clone Detection in Wireless Sensor Networks

Energy and Memory Efficient Clone Detection in Wireless Sensor Networks

ieee projects 2016 for cse, it, eee, ece

Energy and Memory Efficient Clone Detection in Wireless Sensor Networks

ABSTRACT:

In this paper, we propose an energy-efficient location-aware clone detection protocol in densely deployed WSNs, which can guarantee successful clone attack detection and maintain satisfactory network lifetime. Specifically, we exploit the location information of sensors and randomly select witnesses located in a ring area to verify the legitimacy of sensors and to report detected clone attacks. The ring structure facilitates energy-efficient data forwarding along the path towards the witnesses and the sink. We theoretically prove that the proposed protocol can achieve 100 percent clone detection probability with trustful witnesses. We further extend the work by studying the clone detection performance with untrustful witnesses and show that the clone detection probability still approaches 98 percent when 10 percent of witnesses are compromised. Moreover, in most existing clone detection protocols with random witness selection scheme, the required buffer storage of sensors is usually dependent on the node density, i.e., Oð ffiffiffinpÞ, while in our proposed protocol, the required buffer storage of sensors is independent of n but a function of the hop length of the network radius h, i.e., OðhÞ. Extensive simulations demonstrate that our proposed protocol can achieve long network lifetime by effectively distributing the traffic load across the network.

EXISTING SYSTEM:

  • To allow efficient clone detection, usually, a set of nodes are selected, which are called witnesses, to help certify the legitimacy of the nodes in the network. The private information of the source node, i.e., identity and the location information, is shared with witnesses at the stage of witness selection. When any of the nodes in the network wants to transmit data, it first sends the request to the witnesses for legitimacy verification, and witnesses will report a detected attack if the node fails the certification. To achieve successful clone detection, witness selection and legitimacy verification should fulfill two requirements: 1) witnesses should be randomly selected; and 2) at least one of the witnesses can successfully receive all the verification message(s) for clone detection.
  • Randomized Efficient and Distributed protocol (RED) and Line-Select Multicast protocol (LSM) use up their batteries due to the unbalanced energy consumption, and dead sensors may cause network partition, which may further affect the normal operation of WSNs.

DISADVANTAGES OF EXISTING SYSTEM:

  • Is to make it difficult for malicious users eavesdrop the communication between current source node and its witnesses, so that malicious users cannot generate duplicate verification messages.
  • The existing system does not make sure that at least one of the witnesses can check the identity of the sensor nodes to determine whether there is a clone attack or not.
  • Does not guarantee a high clone detection probability, i.e., the probability that clone attacks can be successfully detected, it is critical and challenging to fulfill these requirements in clone detection protocol design.
  • The design criteria of clone detection protocols for sensor networks should not only guarantee the high performance of clone detection probability but also consider the energy and memory efficiency of sensors.
  • The first occurrence of a sensor that runs out of energy, it is critical to not only minimize the energy consumption of each node but also balance the energy consumption among sensors distributively located in different areas of WSNs

PROPOSED SYSTEM:

  • In this paper, besides the clone detection probability, we also consider energy consumption and memory storage in the design of clone detection protocol, i.e., an energy- and memory-efficient distributed clone detection protocol with random witness selection scheme in WSNs.
  • Our protocol is applicable to general densely deployed multi-hop WSNs, where adversaries may compromise and clone sensor nodes to launch attacks.
  • We extend the analytical model by evaluating the required data buffer of ERCD protocol and by including experimental results to support our theoretical analysis. Energy-Efficient Ring Based Clone Detection (ERCD) protocol.
  • We find that the ERCD protocol can balance the energy consumption of sensors at different locations by distributing the witnesses all over WSNs except non-witness rings, i.e., the adjacent rings around the sink, which should not have witnesses.
  • After that, we obtain the optimal number of non-witness rings based on the function of energy consumption.
  • Finally, we derive the expression of the required data buffer by using ERCD protocol, and show that our proposed protocol is scalable because the required buffer storage is dependent on the ring size only.

ADVANTAGES OF PROPOSED SYSTEM:

  • The performance of the ERCD protocol is evaluated in terms of clone detection probability, power consumption, network lifetime, and data buffer capacity.
  • Extensive simulation results demonstrate that our proposed ERCD protocol can achieve superior performance in terms of the clone detection probability and network lifetime with reasonable data buffer capacity.
  • The experiment results demonstrate that the clone detection probability can closely approach 100 percent with untrustful witnesses.
  • By using ERCD protocol, energy consumption of sensors close to the sink has lower traffic of witness selection and legitimacy verification, which helps to balance the uneven energy consumption of data collection.

SYSTEM ARCHITECTURE:

Energy and Memory Efficient Clone Detection

BLOCK DIAGRAM:

block diagram Energy and Memory Efficient Clone Detection in Wireless Sensor Networks

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS: 

  • System : Pentium Dual Core.
  • Hard Disk : 120 GB.
  • Monitor : 15’’ LED
  • Input Devices : Keyboard, Mouse
  • Ram : 1 GB

SOFTWARE REQUIREMENTS:

  • Operating system : Windows 7.
  • Coding Language : JAVA/J2EE
  • Tool : Netbeans 7.2.1
  • Database : MYSQL

REFERENCE:

Zhongming Zheng, Student Member, IEEE, Anfeng Liu, Member, IEEE, Lin X. Cai, Member, IEEE, Zhigang Chen, Member, IEEE, and Xuemin (Sherman) Shen, Fellow, IEEE, “Energy and Memory Efficient Clone Detection in Wireless Sensor Networks”, IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 15, NO. 5, MAY 2016.

Ph.D. Assistance in Chennai

Ph.D. Assistance in Chennai

If you’re looking for assistance with your Ph.D. you’ve come to the right place. JP Infotech is a locally based company located in Chennai that has been assisting students with their Ph.D. and become apart of their success. We provide research and analysis work generating the results you’re looking for. Ph.D. Assistance in Chennai is not a simple task, you’ve already invested your time and money into it, it’s even more beneficial to invest with us so by the end of everything, you know that you’ve maximized resources, getting the best results possible. We’re so confident in our work that we will guarantee you 100% satisfaction.

What We Offer:

Ph.D. Research & Statistics

We have a team of professional Ph.D.’s that can help you along the way. When it comes to research, you might not know what to look for. Our team knows the in’s and out’s about every topic related to Ph.D.’s so we can target the right subject matter with the best possible answers. This is one of the most important steps to doing research, if you don’t know what you’re looking for, you’ll be missing out on a lot of useful information.

Editing Service

It doesn’t hurt to get a second opinion. Without one, it’s very possible to have inaccurate information or even missing key points. With our editing service, we will review your work, correct any information that might be deemed false and also add any relevant information you might’ve forgotten. Our editing service will make you feel confident of your submission prior to deadline. Trust me, you’ll sleep a lot better the night before!

Formatting

Formatting is very important when it comes to professionalism. When reviewing your work, if it’s formatted properly, it will help you qualify for your Ph.D. You want to look like a natural in your field. Let us help review and format your work to make sure you don’t look like an amateur and to help your work stand out from others who are submitting their work. Looks is always important, we will help you properly format your work so yours will not look like the rest.

Complete Ph.D Test

When it comes to test time, how will you react? We have the perfect solution to give you confidence before the test. Our team of experts have compiled a complete Ph.D. test so you will know what to expect. We will train you before it comes to test time, we will ask you many questions that will appear on the time and make sure you know the correct answer. We will make you do it over and over again until you get it right. Our team of experts will grade you and correct any mistakes made to help you better prepare. This is the absolute best solution to give you the confidence you need before it comes to test time.

JP Infotech is one of the top rated companies Chennai when it comes to Ph.D. assistance in chennai. Investing your money with us is equivalent to investing into your future. Knowledge is key and this is what we’re here to help you with. Give us a call with any questions or concerns and we will be more than happy to help. You only get one chance, let us help you build a brighter future.

For Your Ph.D. Research Assistance in Chennai/ Puducherry/ Or from any part of the Globe can directly visit our office or call us to Our Telephone Number or whatsapp or Online etc. The following is our Office address:
Chennai Office: JP INFOTECH, Old No.31, New No.86, 1st Floor, 1st Avenue, Ashok Pillar, Chennai – 83. Landmark: Next to Kotak Mahendra Bank / Bharath Scans.
Landline: (044) – 43012642 / Mobile: (0)9952649690

Pondicherry Office: JP INFOTECH, #37, Kamaraj Salai, Thattanchavady, Puducherry – 9.
Landline: (0413) – 4204066 / (0)9952649690

Email: jpinfotechprojects@gmail.com,
Mobile: (0)9952649690,
Website: http://www.jpinfotech.org,