Literature Review For Domain Generated Algorithms Computer Vs Human

Detecting Algorithmically Generated Malicious Domain Names

Discuss about the Literature Review for Domain Generated Algorithms Computer Vs Human.

This is the literature review of the research project regarding the human generated domain name algorithm and computer generated domain algorithm. This study is about the relationship between the human and computer. In this area, internet is most important factor as it is the main connectivity to the world through the computer technology. Domain name system is the base of the internet technology. In order to use the internet technology efficiently, the domain name algorithm is very important, because the operations of the internet technology is based on the domain name system, of the websites. In this context, the efficiency of the human generated domain name and the computer generated domain algorithm need to be understood. Many researchers have worked in this area in different types of methods. In this literature review, seven different research articles have been reviewed to analyze the two types of domain name generation algorithms.

Yadav et al. (2010), conducted research work regarding the DNS based “domain fluxing”. They have focused in the area of command and control where each of the Bot queries has to register for one domain name only among the available domain names. They have compared the patterns of the domain names of the human generated algorithm and that of the computer generated algorithms. According to them the alphanumeric characters generated in the domain name system are the most important components and need to have the main focus while developing the domain name algorithms. Recent botnets like Torping, Kraken and Conflicker have been analyzed in the research work to understand the malware in the domain name system. In the testing period, the researchers have checked the domain names randomly. In order to conduct the research work in the area of the domain generation algorithms, he understanding of the malicious domain names is the most important fact. This article will help to uncover the area of understanding the malicious domain names by the different types of the domain generation algorithms. Therefore, the review will help to understand the efficiency of the human and computer generated domain generation algorithms for detecting the malicious domains.

The aim of the researcher in this study was to find out the malware, which uses the domain fluxing for circumventing the domain name blacklisting. This strategy is useful for generating the domain generation algorithms. In this research paper, the researcher has said about the procedure for discovering DGAs from Domain Name Service (DNS) query data. To identify certain malware that are responsible for evading blacklist use domain diffusion. There are certain methods used to create an innovative Domain generation algorithm (DGA) to find the domains generated by the use of algorithms. This research is being done to find out the DGA’s from the data query of Domain Name Service (DNS). This system finds out the IP address with distribution which seems unusual of string lengths of second-level. This research paper defines the unidentified DGA’s and the procedures used to find these algorithms. In the following procedure, results of five days of algorithm searching is done after which the final result has been compiled (Mowbray & Hagen, 2014). By finding out the IP addresses of the domain generated algorithms, provides security to the random DNS and IP addresses by giving very less amount of information about the domain names. This article review also helps in traversing the data sets and finding out the unidentified domain names from the given set of DNS so that it can be distinguished from identified domains. Since this algorithm is automated it runs in a time frame with small chunks of time slots. This system also helps in finding the domains in a very large set of data with different dates.

Finding Domain-Generation Algorithms by Looking at Length Distributions

Li et al. (2013), sated about the story generation algorithm for the domain name system. . It is very costly and exhaustive to create such domain sequences manually as they are not scalable. So this system of story creation identifies the correct arrangement even in the unidentified domains to give a narrative story (Li et al., 2013). This system also creates the domain samples from the defined space of the model. A survey of huge number of variables shows that this automated system creates better stories than untrained human writers. This is the future of artificial intelligence in which the system can create infinite number of stories with a given set of names, places, characters and scenarios. This system also includes the Scheherazade System of novel creation with a fictional story on the topic provided by the user. This system uses crowd sourcing which breaks down a complex job into numerous simple tasks. Another method used in Plot Graph learning which is a derivation method of other stories with the reference of which many new stories can be created. The review of this article will help to analyze the efficiency of a particular algorithm in the area of the research topic.

According to Levy (2016), conducted research work on development of a website required different algorithm and also several types of domain. He also gave focus on how these algorithm and domains are affecting the flexibility of the development of proper website by a developer. In this particular research article, it is found that computer language and writing is known as algorithm which is different from the normal words and reading done by the human being generally. It also identified that multitude algorithm is basically available as per the demand but it is not also satisfying the effort given by the human for its development. The automatic narrative generation in the modern technology industry is creating new revaluation (Levy, 2016).  Nowadays companies are using several burden related to financial earnings and projection in the form of algorithm which is narrated in the format of technological language for regulating the mentioned database. The generation of algorithm and natural language generators are also getting complex due to the difference between the artificial call  and the human intelligence. There are several rich domains like finance, sports and reporting related required new software technological platform to demonstrate their advanced patterns, which will help to understand the efficiency between the human and the computer algorithm.

Story Generation with Crowd Sourced Plot Graphs

According to the research article written by Podolny (2015), is initiated the tracking and characterizing the Botnets by using automatically generated domains for creating a particular websites relaying on the modern domain generation algorithms. There are several algorithms which are automatically generated from the domain generation algorithm by focusing on the DNS traffic based on botnets communication capabilities. Generally all the systems has been used in the real world settings, whereas it is supportive for the researchers to measure the intelligence of and created domains which is belongs to botnets automatically. The evaluation and modernization of the 1,153,516 domains are applicable for the labeling the correct botnet simultaneously. There are different approaches which is provided informing that isolated AGD which is helpful for creation of DGA (Podolny, 2015). While development of and domain for a website having several challenges related to DGA modeling, data collection challenges and other lack of grounded facts. According to the overall research study, it is evaluated that improvement in the tracking and monitoring of a DGA and C&C system is required. There are several limitations which is identified and omitted by the phoenix accordingly.  This article review will help to understand the benefits and limitations of the computer generated algorithm, which is very important for the research work in regarding the human generated and computer generated domain name algorithms.

There are many research papers intended to help identify whether an algorithm can make better systematic reporting platform than a human being based on the amount of mistakes. A Chicago based company has been decided to manage the loop for better online product than printing physical articles in the market. The online posted articles are operated on several websites and article publisher and with the support of this company are able to provide better news services and others (Schiavoni et al., 2013). As per the different researchers and media released it is performed that the narrative science category is basically created on the basis of customizing the proprietary artificial intelligence platform for converting the stories and articles. The narrative science is basically related to each and every aspect whether it is related to financial scenario, historical scenario or others. As per the Hammond it is explained that such kind of journals and articles basically used for food chain, financial data sharing and others information sharing activities through the medium of technology. This article is helpful for understanding the efficiency of the computer generated algorithms which is a major section of the current research work.

Can an Algorithm Write a Better News Story Than a Human Reporter?

Extraction of Domain name generation algorithms automatically from the executing malware is another system of security used on the internet. This is done by deploying botnets who not only help in espionage or spamming but also responsible for Distributed Denial of Service attacks. This research paper helps in the classification of different types of DGAs used for the automated extraction of Domain name generation algorithms so that the botnets can be found and eliminated. This system will help to contain the bot headers who control the botnets. This will also help to restrict the domain fluxing and executing the countermeasures such as previously registering the domain names to be used in future and notification can be send to the victims domain names earlier. This research classifies different types of DGAs, it proposes the solution of analysing techniques such as static and dynamic for extraction of the DGAs, it creates an extraction structure for automated DGA (Barabosch et al., 2012). This article review will help in enhancing the security on the internet which can be compromised by the malwares and botnets who use domain names as a weapon to harm the users. Since the number of malwares on internet has increased exponentially this research will enormously help in tracking down the botnets.

Conclusion:

After conducting the literature review regarding the efficiency of the human developed and computer generated domain name algorithms, it can be concluded that the domain name algorithms are very vital for using the internet. The alpha numeric codes used in the domain name system are one of the most vital factors. The main identifying factor between the human generated and computer generated domain name algorithms is the ability of malicious activities detecting. 

References:

Barabosch, T., Wichmann, A., Leder, F., & Gerhards-Padilla, E. (2012, September). Automatic extraction of domain name generation algorithms from current malware. In Proc. NATO Symposium IST-111 on Information Assurance and Cyber Defense, Koblenz, Germany.

Levy, S. (2016). Can an Algorithm Write a Better News Story Than a Human Reporter?. WIRED. Retrieved 1 October 2016, from https://www.wired.com/2012/04/can-an-algorithm-write-a-better-news-story-than-a-human-reporter/

Li, B., Lee-Urban, S., Johnston, G., & Riedl, M. (2013, June). Story Generation with Crowdsourced Plot Graphs. In AAAI.

Mowbray, M., & Hagen, J. (2014, November). Finding domain-generation algorithms by looking at length distribution. In Software Reliability Engineering Workshops (ISSREW), 2014 IEEE International Symposium on(pp. 395-400). IEEE.

Podolny, S. (2015). If an Algorithm Wrote This, How Would You Even Know?. Nytimes.com. Retrieved 1 October 2016, from https://www.nytimes.com/2015/03/08/opinion/sunday/if-an-algorithm-wrote-this-how-would-you-even-know.html?_r=1

Schiavoni, S., Maggi, F., Cavallaro, L., & Zanero, S. (2013). Tracking and characterizing botnets using automatically generated domains. arXiv preprint arXiv:1311.5612.

Yadav, S., Reddy, A. K. K., Reddy, A. L., & Ranjan, S. (2010, November). Detecting algorithmically generated malicious domain names. In Proceedings of the 10th ACM SIGCOMM conference on Internet measurement (pp. 48-61). ACM.