NoyaFangzhou / Athena

Athena

This repository contains the audio evaluation algorithm used in Athena

Swift Version

To import the algorithm, you can simply add the EvaluationBrain.swift file into the directory that contains all the other .swift files Then add following code into your project

var brain = EvaluationBrain()
/* Get the overall score for str1 compared with str2 */
var score = brain.calSimilarity(test: str1, standard: str2)

print(score)

/* Get the detail for str1 compared with str2
 * includes which words in str1 is missing and which words in str1 need practice
 */
var errArr = brain.getErrorArray(test: str1, standard: str2)

for e in errArr {
    print(e.toString())
}

Here str1 is the string to be tested and str2 is the standard string. The return value is a double variable indicates the similarity level

If you still not sure how to run the EvaluationBrain, please see the main.swift file, here I gives a small demo that shows the similarity level between "I sheep" and "I am a ship". You can copy-paste the code into your project.

NPTuple

You can see there is a NPTuple.swift in the EvaluationApp directory. This is a self-designed class storing all strings that need further practice. Basicly, it has consists of 3 parts:

• test string: the string that need further practice, if the test string is empty, this means that the relevant standard string is missing

• standard string: the standard string for the correct pronounciation

• score: the score (used for color)

Also, I designed some functions for this data structure, to make it easy to use. Here is some API

• init(lib: String, test: String, score: Double)
  Constructor for this data structure, you should pass a standard string, test string and a score

• func toString() -> String
  Transfer the current NPTuple to a string that display all its instance variables in a pretty string format

• func needCompare() -> Bool
  This function told whether this is a missing or refine NPTuple, if it return true, it means the user miss the words, otherwise it will return false.

Thanks to NAlexA's answer on StackOverflow and his String extension code, which makes the implementation of this algorithm easier. Here is the link of his/her answer