Homework Help is Here – Start Your Trial Now!
Engineering
Cybersecurity
Word Count Universe of discourse (U): Set of possible word counts (e.g., U = [0, 1000]) Emotional Tone Universe of discourse (U): Set of emotional tone scores (e.g., U = [- 1,1]) Source Reliability Universe of discourse (U): Set of reliability ratings (e.g., U = [0, 1]) "Low (x) = "Medium (x)= 1 200-x 200-1000 0 if x ≤200 if 200x1000 if x 1000 0 x-200 if x 200 or x ≥ 800 if 200x800 = 500 1000-x 200 if 800x1000 1 if x ≤0.2 1 if x 0.5 "Negative (x) = -0.5-x -0.5-0.5 - if 0.5 x < 0.5 "Low (x)= 0.2-x 0.2-0.8 0 if 0.2x0.8 if x ≥ 0.8 0 if x 0.5 0 if x 0.5 or x ≥ 0.5 0 x-0.2 if x 0.2 or x≥ 0.9 "Neutral(x) = x +0.5 if 0.5 x < 0.5 "Medium (x)= if 0.2 < x < 0.9 1 0.7 1-x if 0.9 x 1 0.1 "High h(x) = 0 x-800 200 1 if 800x1000 if x 800 0 if x 0.5 "Positive (x) x-0.5 = if 0.5 < x < 1 0 if x 0.9 0.5 if x 1000 1 if x ≥ 1 Manigh(x)= x-0.9 if 0.9 < x < 1 0.1 1 if x ≥ 1

Word Count Universe of discourse (U): Set of possible word counts (e.g., U = [0, 1000]) Emotional Tone Universe of discourse (U): Set of emotional tone scores (e.g., U = [- 1,1]) Source Reliability Universe of discourse (U): Set of reliability ratings (e.g., U = [0, 1]) "Low (x) = "Medium (x)= 1 200-x 200-1000 0 if x ≤200 if 200x1000 if x 1000 0 x-200 if x 200 or x ≥ 800 if 200x800 = 500 1000-x 200 if 800x1000 1 if x ≤0.2 1 if x 0.5 "Negative (x) = -0.5-x -0.5-0.5 - if 0.5 x < 0.5 "Low (x)= 0.2-x 0.2-0.8 0 if 0.2x0.8 if x ≥ 0.8 0 if x 0.5 0 if x 0.5 or x ≥ 0.5 0 x-0.2 if x 0.2 or x≥ 0.9 "Neutral(x) = x +0.5 if 0.5 x < 0.5 "Medium (x)= if 0.2 < x < 0.9 1 0.7 1-x if 0.9 x 1 0.1 "High h(x) = 0 x-800 200 1 if 800x1000 if x 800 0 if x 0.5 "Positive (x) x-0.5 = if 0.5 < x < 1 0 if x 0.9 0.5 if x 1000 1 if x ≥ 1 Manigh(x)= x-0.9 if 0.9 < x < 1 0.1 1 if x ≥ 1

icon
Related questions
Question

Please dont USE Ai 

Fuzzy Logic-Based Fake News Detection Example

Suppose we have a news article with the following features that you use to determine whether the article is likely to be fake news:

  • Word Count: The total number of words in the article.
  • Emotional Tone: A score indicating the article's emotional tone (positive, negative, or neutral).
  • Source Reliability: A rating (on a scale of 0 to 1) representing how reliable the news source is.

To describe the qualitative characteristics of the above features and represent its inherently vague or imprecise, we use these linguistic variables:

  • Word Count: Low, Medium, High
  • Emotional Tone: Negative, Neutral, Positive
  • Source Reliability: Low, Medium, High

These membership functions represent each feature.

     
     

Consulting a group of experts, they suggest that to classify an article as fake news, we must follow two basic rules:

  • If the Word Count is High AND the Emotional Tone is Negative, the article is likely fake.
  • If Source Reliability is Low, then the article is likely fake.

The final step is the defuzzification process. It is based on a threshold defined for the problem. In the current process, the crips value resulting from the defuzzification process is more significant than 0.7, and then the article will be considered fake news.

We use a MAX aggregation function and a centroid method (COG) to find this crisp value. The defuzzied value (COG) can be calculated using this formula.

Where:

  • µi the degree of membership for the i-th fuzzy set.
  • xi​ is the representative value (often the centroid) of the iii-th fuzzy set.

 

Using the previous model, calculate whether the following articles could be considered fake. Answer if the article is fake or not!

Demonstrating the calculus is required.

Article 1

Word Count: 600, Emotional Tone: 0.4, and Source Reliability: 0.6

 

Fuzzification Process:    Participants demonstrate the calculation to infer the fuzzy value through the crisp one’s existence in the problem formulation.

 

Defuzzification Process:              Participants demonstrate the calculation to infer the crisp value through the aggregation and application of the COG function.

 

Calculations:     The calculations do not have any mistakes (wrong calculus).

              The calculations are precise (the equations are correctly applied).

 

Result:  The Participants cites if the paper is fake or not!

Word Count Universe of discourse (U): Set of possible word counts (e.g., U = [0, 1000]) Emotional Tone Universe of discourse (U): Set of emotional tone scores (e.g., U = [- 1,1]) Source Reliability Universe of discourse (U): Set of reliability ratings (e.g., U = [0, 1]) "Low (x) = "Medium (x)= 1 200-x 200-1000 0 if x ≤200 if 200x1000 if x 1000 0 x-200 if x 200 or x ≥ 800 if 200x800 = 500 1000-x 200 if 800x1000 1 if x ≤0.2 1 if x 0.5 "Negative (x) = -0.5-x -0.5-0.5 - if 0.5 x < 0.5 "Low (x)= 0.2-x 0.2-0.8 0 if 0.2x0.8 if x ≥ 0.8 0 if x 0.5 0 if x 0.5 or x ≥ 0.5 0 x-0.2 if x 0.2 or x≥ 0.9 "Neutral(x) = x +0.5 if 0.5 x < 0.5 "Medium (x)= if 0.2 < x < 0.9 1 0.7 1-x if 0.9 x 1 0.1 "High h(x) = 0 x-800 200 1 if 800x1000 if x 800 0 if x 0.5 "Positive (x) x-0.5 = if 0.5 < x < 1 0 if x 0.9 0.5 if x 1000 1 if x ≥ 1 Manigh(x)= x-0.9 if 0.9 < x < 1 0.1 1 if x ≥ 1
Transcribed Image Text:Word Count Universe of discourse (U): Set of possible word counts (e.g., U = [0, 1000]) Emotional Tone Universe of discourse (U): Set of emotional tone scores (e.g., U = [- 1,1]) Source Reliability Universe of discourse (U): Set of reliability ratings (e.g., U = [0, 1]) "Low (x) = "Medium (x)= 1 200-x 200-1000 0 if x ≤200 if 200x1000 if x 1000 0 x-200 if x 200 or x ≥ 800 if 200x800 = 500 1000-x 200 if 800x1000 1 if x ≤0.2 1 if x 0.5 "Negative (x) = -0.5-x -0.5-0.5 - if 0.5 x < 0.5 "Low (x)= 0.2-x 0.2-0.8 0 if 0.2x0.8 if x ≥ 0.8 0 if x 0.5 0 if x 0.5 or x ≥ 0.5 0 x-0.2 if x 0.2 or x≥ 0.9 "Neutral(x) = x +0.5 if 0.5 x < 0.5 "Medium (x)= if 0.2 < x < 0.9 1 0.7 1-x if 0.9 x 1 0.1 "High h(x) = 0 x-800 200 1 if 800x1000 if x 800 0 if x 0.5 "Positive (x) x-0.5 = if 0.5 < x < 1 0 if x 0.9 0.5 if x 1000 1 if x ≥ 1 Manigh(x)= x-0.9 if 0.9 < x < 1 0.1 1 if x ≥ 1
Expert Solution
steps

Step by step

Solved in 2 steps

SEE SOLUTIONCheck out a sample Q&A here
Blurred answer