Designing sounds and building sound landscapes is a challenging yet inspiring occupation. Dealing with large data sets and having numerous influencing factors are some of the concerns making the situation more demanding. Utilizing sound libraries containing SFX is one way to make the process more approachable and productive. However, many aspects still need improvement to make the means of working with the libraries more gratifying and compelling. No standard exists for categorizing and classifying sounds, meaning that most libraries have a custom labeling system, making it harder to develop programs capable of handling all variations. The tools used for finding and managing sounds among the libraries are also far from perfect. Extensive research has taken place in automatic sound identification and objective audio labeling, which could help classify sounds more accurately and expressly. The research results and developed systems thereof could assist the library creators in the tagging process, making related sounds across multiple libraries labeled similarly. This procedure could aid in establishing a standard of how to describe sounds concretely, such as the sound source of the audio. Far less experimentation has ensued with systems aimed towards individual sound annotation with personal descriptors, capable of adjusting to different users' perception. Humans often describe signals perceived through the senses with subjective words, e.g., that something is scary. This impression holds for audio as well: one person might interpret a sound as 'happy' when another defines it as 'smooth.' The purpose of this thesis is to present and evaluate an approach capable of categorizing sounds with subjective tags. The developed solution is a flexible stand-application application, which continuously improves with usage and minimizes the manual effort needed to label sounds utilizing a sound similarity and automatic tagging system. Although the program functions well on its own, the eventual intent is to have the system integrated with existing categorization methods and management tools. This procedure could hopefully improve the work process for sound designers using sound libraries and also spark ideas for new programs or extensions to systems regarding the subject.