• performance
• co-occurence analysis
• startups
• business accelerators
• business incubators

It is well known that only a low percentage of start-up companies survives on the market and even less become large successful companies. Business incubators and accelerators are programs designed to improve start-up development through a variety of business support services. The thesis examines the literature on incubators and accelerators using text mining techniques on a sample of 1048 documents. A keywords analysis is performed using TF-IDF and co-occurrence and I identified the most important topics. Despite the large literature, there is still considerable uncertainty about the factors that explain the performance of these programs. The thesis carries out a detailed examination of the empirical findings in literature and it identifies two main groups of antecedents: first, nature and quality of services offered by programs such as funding, infrastructure, mentoring and networking and second, internal factors, in particularly the governance of the institutions, and external factors (surrounding environment or ecosystem). Finally, I identified an inverted U-shape relation between the number of start-ups enrolled in the programs and performance of incubators and accelerators.