Generalized Vector Space Model Implementation System In Web-Based Search Song Lyrics
By : Km Ressa Caprytenta Trismana
Faculties : Fakultas Matematika dan Ilmu Pengetahuan Alam
Department : S1 Ilmu Komputer
Ease of information exchange becomes very important in all aspects of life. At the moment, most people like the songs of many kinds of songs of different genres such as the flow of pop, rock, jazz and others. Because the times are already global people easily obtain songs through the Internet and people want to sing the song also look for the lyrics on the Internet. Problems often faced by users when they search the lyrics of the song that is when the user forgot the title or name of the singer who wants to be sought by the user, in other words the user must know a few words of the lyrics of the song you want to search by the user. If a user is given a lot of words in the lyrics, the possibility of the lyrics is exactly what will appear the more words the user inputted the chances of the lyrics will appear. And so the problem can be achieved required exact algorithm in solving the problems that the algorithm Generalized Vector Space Model to calculate the frequency of the relevant lyrics. Generalized vector space model is the development of vector space model by conducting inter-term considerations more accurate sense. With this method increases the recall of the search results, so that the search results include more relevant documents. The average value of which is derived from the use of methods GVSM has an average of 86.87% precision and recall the average value of a high of 93.74%. This shows the system is able to produce search lyrics that match the input query entered by the user.
Keyword : Information Retrieval, Generalized Vector Space Model, Preprocessing, Stemming, Precision, Recall, Term