OBJECT ORIENTED METRIC BASED ANALYSIS OF TEXT TRANSMISSION IN E-LEARNING THROUGH NATURAL LANGUAGE STEGANOGRAPHY

E-learning is an application of information and communication technology in the field of learning. Through steganography the e-learning institution can provide security to other participants of e-learning like teacher and learner. Here we use text steganography with modified SNOW algorithm while passing secret texts from the administrator to the learner in an e-learning system. In this paper, we calculate the object oriented metric based analysis of CK and MOOD metrics of our proposed model, which ensures the advantages of code redundancy, code reusability, and cost effectiveness and so on.


Introduction
Security plays a great role in any efficient e-learning system which should be provided by the administrator of the e-learning institution [1] . To make log in into any system, two essential requirements are log-in id and password and both of these are generally text based documents and very important to keep secret for security purpose. If these documents move to wrong hand then it will be very harmful for the e-learning institutions. Here we present our model regarding the transmission of texts securely from the administrator to the teacher in an e-learning system. Now, if an outsider knows about the log-in id and password to enter into an e-learning system, then he can easily upload wrong study material, can download all the related documents of study which was provided by that teacher, can upload wrong question papers and so on, which not only become a threat for the teacher but also make a bad reputation for the institution and also become harmful for the student. So the transmission of these texts should be kept secret. Text steganography is a technique through which the sender can hide the secret texts covered into another text file while sending to the receiver [2] . To provide security, here we have used a modified version of Steganographic Nature of Whitespace [3] , where we apply Advanced Encryption Standard (AES) as a replacement of Information Concealment Engine (ICE) algorithm [4] . The whole process can be divided by three parts: compression, encryption and encoding scheme. Here we apply Huffman coding, which is a lossless data compression algorithm and the main two applications are to build Huffman Tree from input characters and traverse the Huffman Tree and assign codes to characters [5] .
Then we apply AES algorithm for the encryption of 128 bit key which is sufficient to protect information at a secret level [6] and also easy to implement with a high speed and low RAM.
At the beginning of the message, immediately append after the text after the first line to separate the block of spaces, which is not possible unless the last 3 bits coded to zero spaces.
Metrics are units of measurement and according to E.V. Berard the five main characteristics of object oriented metric based analysis [7] are Localization, Encapsulation, Information hiding, Abstraction and Inheritance [8] . The advantages of object oriented metric based analysis over traditional system are reduction of maintenance cost, reuse of code, improvement of portability, compatible with real world system and so on. Here we calculate the values of object oriented metrics related to Chidamber and Kemerer (CK) Metrics and Metrics for Object Oriented Design (MOOD) [9] , which are mostly used in metric analysis.
In this paper, we have analyzed the values of the object oriented metrics based on the class hierarchy diagram of the proposed model regarding the transmission of texts securely from the developer to the teacher. Section II includes the class hierarchy diagram of the proposed model regarding the transmission of texts secretly from the developer to the teacher in an e-learning system. Section III covers the metric based analysis of the proposed model based on the class diagram and finally, we conclude in section IV by showing some future scope.

Class Diagram of Proposed Model
Class diagram is a part of the static structural Unified Modeling Language (UML) diagram, which is used to represent the structure of a system by showing the classes of the system and their attributes, operations (or methods) and the relationship among objects [10] . The class diagram of our proposed model is shown in Figure1 [11] regarding the secure transmission of text from developer to learner. In this class diagram we have used five classes: HuffmanTree, HuffmanEncoder, AESencryption, Node and StegoMain. [70]

Metric Based Analysis of Proposed Model
In this paper, we calculate the values of the Chidamber and Kemerer (CK) metrics and Metric for Object Oriented Design (MOOD) metrics [12] . Some metrics of the CK metrics are discussed below:   Now, we will graphically represent the above values and discuss on these:  Here we can see the maximum value is 12, which indicates that the system is easy to maintain.
where A d (C i )= total attributes defined in class C, A v (C i )= Attributes visible in class C and A h (C i )= attributes hidden in class C. The AHF value of our proposed model is 0.146, shown in Table2 that lies in between 0 and 1, which is acceptable.
[73]  The value of MHF of our proposed model is calculated by using Table3, which is 0.048. Low value of MHF means insufficiently abstracted implementation and high MHF indicates the less functionality. So, this value is okay [14] .  The value of AIF is 0.672, that is shown using Table4. If the value of AIF is 0, then it means lack of inheritance but here the value is greater than 0 and less than 1, which is acceptable.