The vast growth of information and communication technology resulted in a hugernvolume of information very large bulk of which is stored as unstructured text. Thernpresence of so much text in electronic form is a challenge to natural languagernprocessing As the volume of electronic information increases, there is growing interestrnin developing tools to help people better find , filter, and manage these resources.rnArguably, the only way for humans to cope with the information explosion is to exploitrncomputational techniques that can sift through huge bodies of text.rnCurrently news agencies in Ethiopia in which large amount of news from all the availablernsources are processed every day is implementing a manual classification system torncategorize news items in their daily activities despite the fact, they are usingrncomputerized system to store and edit news items. Radio Fana is the one among thesernagencies.rnThe objective of this research is to develop and adopt processing tools for Afaan Oromorntext classification and investigate the application of machine learning techniques forrnautomatic classification of Afaan Oromo news items.rnThe data source for this research is the Afaan Oromo news items obtained from RadiornFauna Share Company.rnIn this research , tools for pre-processing Afaan Oromo news items such as tokenization,rnremoval of extraneous characters, removal of stop-words and removal of affixes fromrnthe words are prepared to facilitate the experimentation process for the automaticrnclassifiers.rnAmong the automatic classifiers which are applicable on high dimensional data, four ofrnthem; Sequential Minimal Optimization (SMO) algorithm from Support Vector Machines,rnNaÃ¯ve Bayes Multi Nominal (NBM) from Bayesian Classifiers, J48 algorithm from thernDecision trees and K-Nearest Neighbor (KNN) from the Lazy Learners have beenrnexperimented on the final data. The data, the pre-processed Afaan Oromo news items,rnis organized in to categories of four classes, seven classes and all (eleven) classes forrnthe experimentation purpose and the experimentation uses 10-fold stratified crossrnvalidation for training and test data.