Letters written by great people in history are circulating on social networks. An example is the letter that the German physicist Albert Einstein (1879-1955) wrote to his daughter Lieserl Einstein. In this letter, he shows that love is the only answer to the survival of our species. There is also a letter that Simon Bolívar (1783-1830) claims to have written to his cousin Fanny du Villars, written on her deathbed, telling her how important she was in his life.

In fact, Albert Einstein's letter to his daughter analyzes the concept of love as a universal force that science has not yet fully developed. This letter says that love is seen as light, gravity and life. It is the only energy in the universe that humans have not learned to control at will. Suggests that love is the most powerful force because it has no limits.

Simón Bolívar wrote a letter to his cousin Fanny du Villars, obtained from the archives of the National Academy of Medicine of Colombia (https://www.interacademies.org), to express his emotional affection for her. In the letter, he describes the surrounding nature and emphasizes the beauty of the Caribbean Sea, the magnificent mountains and the colorful sky, and tells her how important she was in his life. The strange thing is that the real name of the French woman was Louise Jeanne Nicole Arnalde "Fanny" Denis de Keredern de Trobriand (1775-1859).

On the other hand, all the papers of Albert Einstein are collected in various databases such as The Collect Papers of Albert Einstein (https://einsteinpapers.press.princeton.edu/), Einstein Papers Project (https://www.einstein.caltech.edu/), and so on. None of them have any proof of the existence of this letter to his daughter.

The same applies to the documents of Simón Bolívar available at Biblioteca Virtual Miguel de Cervantes (https://www.cervantesvirtual.com). There is also no information in these files about the existence of this emotional letter. For all this, these letters are analyzed according to the stylometric methodology as described below.

Stylometric has been summarized as a statistical methodology based on the frequency analysis of each author's own words 14611. The term was coined by the Polish writer Wincenty 8 when he determined the chronology of Plato's dialogues 8 and was originally based on a chi-square analysis of word length-word and their frequency 9.

More recently, it has been used to determine the authorship of some unknown or questionable works, such as a comedy originally attributed to Miguel Bermúdez by the National Library of Spain, but found by this method to be the author of Lope de Vega. 3.

Another example was Ulysses, published in 1922 by the Irishman James Joyce, but stylometric studies conclude that it was written by five other people who were not mentioned in the work. 12.

Based on the above, the paper analyzes eight letters written by Albert Einstein and finds out if the letter to his daughter was really written by him. This procedure is repeated in a letter that Simón Bolívar allegedly wrote to his cousin, as described below.

Daelemans 4 describes the methodology for performing a stylometric analysis and is summarized below. A corpus is created from the letters of the same author. The selected letters and the numbers were chosen at random (details in Table 1).

For each author, the word frequency is determined according to the Delta function, which is a linguistic measure capable of distinguishing the authorship of texts according to the definition proposed by 2.

Some articles questioned the use of the Delta function, but Burrows showed that it is an excellent technique for identifying the authorship 13. This function simply determines the frequency variation of the most frequent words in the text using z-scores 13. Remember that z-scores are a measure of relative frequency difference minus the word mean divided by the standard deviation 16.

After that, a matrix was calculated based on the distance obtained from the frequency of the words 15. This calculation uses one, two or more words, and this number is indicated as n-grams 15.

Based on this frequency, the distance Remember that there were two possible element types used to generate n-grams: words and characters. Character n-grams show how frequently certain letters, capital letters, punctuation marks, or numbers are used at the alphabetic level of a language, while the word n-grams and vocabulary in a document are connected. In addition to word frequency, these characteristics also include sentence length, word length distribution, richness of vocabulary, and lexical mistakes. These can be used as the initial tokenization step for any language 10.

For other part, it is usually calculated using the Manhattan, Euclidean, and so on 15. The result is usually visualized with a dendogram, a tree where similar distances are grouped based on a certain number of words into a single conglomerate or cluster, simply abbreviated as MRW (ie., Meaning Most Words).

The paper considers that sentences written by the same person should be grouped into the same branch or node 5. All calculations were performed in the Python programming language.

Figure 1 shows the result of normalization of the frequency of words obtained from the English letters of Albert Einstein when the n-gram is equal to two (ie two words). It can be seen that the alleged letter to daughter does not follow the same pattern as her other letters, with the words with the biggest difference such as “of the”, “it is”, “in a”, “is the”, “and the”, “to me”, among others.

Figure 2 shows the result of the normalization of the frequencies of words obtained from the letters of Simón Bolívar using n-gram equal to 1 and MRW = 50. This graph shows differences between the letters, such as “mi”, “los”, “para”, ”he”, “si”, “dinero”, among others.

Figure 3 shows a rectangular dendogram of Albert Einstein's letters. It was observed that the supposed letter to the daughter does not reproduce the linguistic styles of the other eight letters, that is, the supposed letter (labeled Daughter) does not belong to any of the three groups (clusters) that make up the rest of letters.

Figure 4 shows a rectangular dendogram of Simón Bolívar's letters, showing that his cousin Fanny's letter does not belong to any of the two large conglomerates that repeat the style of Bolivar. Therefore, Fanny's letter was not written by Simón Bolívar.

The paper's goal was to determine whether Albert Einstein and Simon Bolivar actually penned amorous letters to their daughter and cousins, respectively, by looking for linguistic evidence in their letters. It is established that they did not write these tearful letters based on the findings of a stylometric examination.