results.rb

gistfile1.rb

results = 
  {
    '123' => [
      1,
      2,
      3
    ],
    '234' => [
      1,
      2
    ],
    '345' => [
      1,
      2
    ],
    '456' => [
      4,
      5
    ],
    '567' => [
      6,
      7
    ]
  }

# Initialize hash of empty arrays
duplicates = Hash.new { |k,v| k[v] = [] }

# If it has more than half the elements, log the result.
threshold = 0.5;

results.each do |primary_id, primary_result|

  # For every item in set check intersection with all other items.
  # If length of intersection is close to length of result,
  # log secondary item.

  results.each do |secondary_id, secondary_result|

    if primary_id != secondary_id
      intersect = primary_result & secondary_result

      if (intersect.length.to_f / primary_result.length.to_f) > threshold
        duplicates[primary_id] << secondary_id
      end

    end

  end

end

puts duplicates