Semantic similarity on constraints datasets: A latent approach

The technological world has grown by incorporating billions of small sensing devices, collecting and sharing large amounts of diversified data over the new generation of wireless and mobile networks. Semantic similarity models have been used as a means to organize and optimize devices in constrained environments such as IoT, edge computing, and 5G and next-generation networks. In this paper, we reviewed the commonly used semantic similarity models, discussed the limitations of our previous model, and explored latent space methods (through matrix factorization) as a way to reduce noise and correct the model profiles with no additional data. Our solution was evaluated on two datasets: Miller-Charles and IoT semantic datasets. The improved model achieved a correlation of 0.62 and 0.53 respectively (which represents an improvement of 0.21 and 0.13 for each dataset).