Knowledge Discovery in Database (KDD) is a structured analysis process aimed at getting new and correct information, finding patterns from complex data, and being useful. Data mining is at the core of the KDD process. Clustering is a data mining method that is suitable for optimizing library services because it can cluster books effectively and efficiently, with the K-Means algorithm data can be clustered and information from each centroid value of each cluster. Library services can optimize the placement of books so that students can quickly find books according to their reading interest more effectively and can be attracted to other books because they are in one grouping. Meanwhile, the library can prioritize the procurement of the next book. Optimization of library services in the cluster using the K-Means method. Clustering interest in reading has the criteria for the number of books available, borrowed books, and the length of time the books are borrowed. The book data is clustered into 3, namely very interested, in demand, and less desirable. After doing the calculation process from 40 samples of book types, it resulted in 6 iterations, and the final results were 3 clustering, namely cluster 1 of 4 books that were of great interest, cluster 2 of 20 books that were of interest, and cluster 3 of 16 books that were less desirable. This research can be used as a recommendation reference for optimizing library services both for the layout and procurement of books by prioritizing the types of books that are of great interest.

Sistem pendukung keputusan merupakan bagian dari sistem informasi berbasis komputer untuk mendukung pengambilan keputusan dalam sebuah organisasi atau perusahaan. Tujuan dari penelitian ini adalah membangun sebuah sistem yang dapat mempermudah sebuah instansi seperti Bimbel Gama Bukittinggi untuk menyeseleksi guru yang layak untuk diangkat menjadi guru tetap yang ditampilkan dalam perangkat lunak aplikasi berbasis sistem pengambilan keputusan, dengan menggunakan metode penggabungan Analytic Hierarchy Process dan Weighted Product. Pada tahap awal sistem ini menggunakan metode Analytic Hierarchy Process untuk menentukan bobot setiap kriteria dan pada tahap berikutnya menggunakan metode Weighted Product untuk perangkingan setiap alternatif, output yang dihasilkan yaitu rekomendasi sebagai guru tetap. Sistem ini dibangun dengan menggunakan program PHP, MySQL berbasis web.