we need matlab code to detect a peson male or female using image processing..
Posts: 14,118
Threads: 61
Joined: Oct 2014
There are well-known linguistic differences between the writing of men and women, and these differences can be effectively used to predict the gender of the author of a document. Taking advantage of these linguistic nuances, this study uses a set of stylometric features and a set of word counting functions to facilitate automatic gender discrimination in emails from Enron's popular e-mail dataset. These features are used in conjunction with the Winnow Balanced Modified Neural Network proposed by Carvalho and Cohen, an improvement on the original balanced Winnow created by Littlestone.
It is widely accepted that there are significant linguistic differences between men and women. Previous studies have analyzed the forms of these linguistic distinctions and their links to social roles. Multiple linguistic characteristics have been determined, such as the use of characters, writing syntax, functional words, and word frequency. Other features, such as those contained in the Psycholinguistics database of the Media Research Center (MRC) and the Linguistic Research and Word Count software (LIWC) have also been examined. Some studies have also divided the text into n-grams for analysis. N-grams are combinations of n words or characters used in sequence to capture the structure of an author's unique writing style.
It has been reported that women tend to use more emotionally charged language, as well as more adjectives and adverbs, and apologize more frequently than men. On the other hand, men use more references to quantity and commit a greater number of grammatical errors. Gender-specific language has also been observed to be more prevalent in single-gender conversations compared to pairs or groups of both sexes. Using the various methods and characteristics, researchers have automated the prediction of an author's genre with accuracies ranging from 80% to 90%. The accuracy expected is often limited by the type of text being evaluated, with certain types of documents proving more difficult than others. For example, business emails have less gender-preferential language than blogs, so commercial emails are harder to classify than blogging and reduce expected accuracy.