Binary search alternatifi algoritma geliştirme üzerine bir araştırma

Abstract

ÖZET Günümüzde yardımcı bellekler üzerine büyük miktarda bilgi toplamak mümkündür. Bu kadar büyük miktarda depolanmış bilgi içerisinden, istenilen bilgiye kısa sürede erişilerek geri alınması son derece önemlidir. Bu araştırmada, hızlı arama yaptığı kabul edilen Binary Search algoritmasına alternatif olabilecek yeni bir algoritma geliştirmeye çalışılmıştır. Araştırmamızın esasını bilgisayarlarda ön bellek ` cache memory` olarak kullanılan statik RAM'ların yüksek erişim hızından yararlanarak, anahtar dosyadan bir defa okuma yaparak arama süresini azaltmak oluşturmaktadır. Bu amaçla, ön belleklerin çalışma prensibi, aramada kullanılacak anahtarın -/ adrese çevrilmesi, disk-dosya işlemleri için handle tekniği incelenmiştir. Ayrıca alternatif algoritmamıza destek vermesi için 64K'lık bir ön bellek arabirim kartı yapılmıştır.II SUMMARY We can store great many data on the auxiliary memories to day. It is important to get access and retrieve the required information from among the previously stored information in a great quantity. In this research, I tried to develop a new algorithm that is an alternative to Binary Search algorithm which is considered to search quickly. The principle of our research is based on decreasing the search time reading an access area from the key file making use of the high access speed of the static RAMs which are used as cache memory in computers. We aimed at studying the working principle of cache memories and converting the key used in searching into address. The handle technique has been studied for disk-file operations. In addition, to support to our algorithm an interface board by 64K cache memory was constructed.

II SUMMARY We can store great many data on the auxiliary memories to day. It is important to get access and retrieve the required information from among the previously stored information in a great quantity. In this research, I tried to develop a new algorithm that is an alternative to Binary Search algorithm which is considered to search quickly. The principle of our research is based on decreasing the search time reading an access area from the key file making use of the high access speed of the static RAMs which are used as cache memory in computers. We aimed at studying the working principle of cache memories and converting the key used in searching into address. The handle technique has been studied for disk-file operations. In addition, to support to our algorithm an interface board by 64K cache memory was constructed.