KLASIFIKASI JENIS KELAMIN PENULIS MENGGUNAKAN ALGORITMA K-NEAREST NEIGHBOUR DAN HARMONY SEARCH; AUTHORS’ GENDER CLASSIFICATION USING K-NEAREST NEIGHBOUR AND HARMONY SEARCH ALGORITHMS

Erick, Aina Musdholifah

2015 | Skripsi | FMIPA

Author profiling is a task to determine profile of the author of some specific documents. Where as profile of an author may consist of some or all of the author’s characteristics such as gender, age, nationality, or personality. There are a lot of algorithms that can be used in implementing author profiling. One of them is k-Nearest Neighbour (k-NN). In k-NN, the test data is classified to the majority class of the k datas that have the highest similarity with the test data. As the numbers of features that are not in a small amount, it may lead to the curse of dimensionality problem in k-NN, Harmony Search Algorithm is used in order to implement feature selection so that the number of features can be optimalized. From various testing scheme through the system, authors’ profile classification using k-Nearest Neighbour and Harmony Search Algorithm yields accuracy with the maximum value at 61% with recall and precision each at the value of 54% and 62.79%.

Kata Kunci : author profiling; k-NN; HSA