This work presents an analysis between two approaches for inferring trust in social networks: TidalTrust and T-SWEETS. The first part of this research is composed by an experiment with the Epinions, Slashdot and Wikipedia datasets in order to define the most suitable type of trust value (integer or fractioned, positive or negative, and so on). Furthermore, the second part of this research reports a comparative analysis between both algorithms. The results indicate that T-SWEETS outperforms TidalTrust in terms of accuracy and maintains the transitivity principle, which is the basic principle of trust.f trust.