Homograph Detector

Based on IDN homograph attacks [1], a program to find all such domain names in a given set 'D' is implemented. Homoglyph characters are identified by using edit distance.

Note: Hamming Distance can also be used to solve this problem.

Data Structure

Adjacent Graph - To find nearby domain based on the edit distance.

Assumptions

• set 'D' consists of valid domain names.

Packages

• default: Has main method.
• domain: Has AdjancentGraph class to load and perform domain operations.

Class

• Main: gets domain set 'D' from user and performs desired operations based on the choice.
• AdjancentGraph: loads and compute edit distance, edit path and nearbyNum.

Methods

• load #domain.AdjancentGraph:

  • Gets the set of domain name 'D'
  • Constructs a @graph as <key, edges[]> which will be in form of adjacency list.
  • Edges are connected between domain which has unit edit distance.

• doGraphTravesal #domain.AdjancentGraph:

  intializes the path, level and other required factors & calls traverseAllPath method.

• traverseAllPath #domain.AdjancentGraph:

  traverse the graph from source to destination recursively by using the adjancent list in the value field of @graph.

• editDistance #domain.AdjancentGraph:

  Gives the smallest number of edits to convert domainName1 to domainName2

• editPath #domain.AdjancentGraph:

  Outputs all possible sequence to convert domainName1 to domainName2 in the given domain set 'D'.

• numNearby #domain.AdjancentGraph:

  Gives the total number of possible by having a max limit set in the given domain set 'D'.

• traverseBydistance #domain.AdjancentGraph:

  traverses the graph to till the specified distance in all directions.

• isUnitDistance #domain.AdjancentGraph:

  Check whether given strings has edit distance of 1 by either inserting or deleting a char.

Reference

[1] Umawing, J. (2018, January 22). Out of character: Homograph attacks explained. Retrieved from https://blog.malwarebytes.com/101/2017/10/out-of-character-homograph-attacks-explained/