Analisis Perbandingan Algoritma (Left-Corner-Parsing Dan Algoritma CYK (Cocke-Younger-Kasami)Untuk Memeriksa Pola Kalimat Baku Bahasa Indonesia

Seorang penulis harus memperhatikan pola kalimat dalam bahasa tulisan, karena pola kalimat akan berpengaruh dalam proses penyampaian dan penerima pesan. Berdasarkan pola dasarnya, bahasa indonesia mengenal beberapa pola kalimat (1) S-P, (2) S-P-O, (3) S-P-Pel, (4) S-P-K, (5) S-P-O-Pel, (6) S-P-O-Pel-K, (7) S-P-O-K, dan (8) S-P-Pel-K. Hal inilah yang mendasari proses penulisan perlu memperhatikan pola kalimat, sehingga perlu diadakan penelitian tentang pemeriksaan pola kalimat.Algoritma yang digunakan dalam memeriksa kalimat sangat banyak macamnya, dua diantaranya algoritma LCP (Left Corner Parsing) dan algoritma CYK (Cocke-Younger-Kasami). Algoritma LCP (Left Corner Parsing) merupakan algoritma gabungan dari Top down Parsing dan Bottom up Parsing. Proses dari algoritma LCP (Left Corner Parsing) dimulai secara Bottom Up dan diakhiri secara Top Down. Algoritma CYK menggunakan struktur array dua dimensi dalam pengecekkannya. Setiap kata diperiksa kelompok katanya, kemudian dilakukan pengecekkan struktur kalimatnya. Dari perbedaan cara kerja algoritma, maka akan dipelajari dan dibandingkan kemampuan masing-masing algoritma dalam memeriksa pola kalimat baku bahasa Indonesia. Kedua algoritma tersebut dilakukan analisis sehingga diketahui tingkat akurasi dari masing-masing algoritma dalam memeriksa pola kalimat baku.Hasil pengujian pemeriksaan pola kalimat baku didapatkan tingkat akurasi dari algoritma CYK (Cocke-Younger-Kasami) sebesar 65% dengan laju error 0.35 dan algoritma left corner parsing sebesar 60% dengan laju error 0.40. Namun tingkat akurasi yang didapat belum cukup besar karena kamus POS Tag yang digunakan masih terdapat kekurangan yaitu masih banyak kata yang tidak dapat dikenali.

An author must pay attention to the pattern of the sentences in the language of the inscription, because the pattern of sentence will be influential in the process of delivery and recipient of the message. Based on the patterns of nature, get to know some indonesian sentence patterns (1) S-P, (2) S-P-O, (3) S-P-Pel, (4) S-P-K, (5) S-P-O-Pel, (6) S-P-O-Pel-K, (7) S-P-O-K, and (8) S-P-Pel-K. This is the underlying process of writing needs to pay attention to the pattern of sentence, so the need to research about the examination pattern of sentence.The algorithms used in checking a sentence very many kinds, two of which are algorithms LCP (Left-Corner-Parsing) and algorithms CYK (Cocke-Younger-Kasami). The algorithm is the algorithm for parsing left corner combination of Top down Parsing and Bottom up Parsing. The process of the algorithm's Left Corner in the Bottom Up Parsing starts and ends in the Top Down. CYK algorithm using two dimensional array structure in examination. Any checked group said, then carried out his sentence structure examination. From the difference of the workings of the algorithm, it will be studied and compared the ability of each algorithm in examining raw language sentence patterns of Indonesia. Both of these algorithms is done so that the analysis of known accuracy from each of the algorithms in the check out raw phrase patterns The test results obtained by examination of raw sentence patterns of degree of accuracy of the CYK algorithm (Cocke-Younger-Kasami) by 65% with an error rate of 0.35 and algorithms left corner parsing by 60% with an error rate of 0.40. But the level of accuracy obtained is not yet big enough because the dictionary POST Tags used there is still a shortage that is still a lot of words that could not be identified.