| dc.description.abstract | Information retrieval system was developed using various models, such as probabilistic models, language models, boolean models, vector-space models and many more. Thus, it’s problematic to determine which models is the best and the most efficient in every search condition. In this study, two models were developed and compared: probabilistic model and vector-space model. The probabilistic model has Okapi BM25 similarity function with parameters that are subject to fine tuning to seek for better performance. Fine tuning the parameters has made the probabilistic model’s average precision increases from 0.5885 to 0.5901. Further, this model also outperformed the vector-space model with average precision 0.5327. | en |
