Tapioca

The Web of data is growing continuously with respect to both the size and number of the datasets published. Porting these datasets to five-star Linked Data however requires data publishers to link their novel dataset with the already available Linked Data sets. Given the size and growth of the Linked Data Cloud, the current mostly manual approach used for detecting relevant datasets for linking is thus obsolete.

We present Tapioca, a linked dataset search engine so as to provide data publishers with similar existing datasets automatically. Our search engine uses a novel approach for determining the topical similarity of datasets. This approach relies on probabilistic topic modelling to determine related datasets by relying solely on the metadata of datasets.

The source code can be found at Github. The software is provided under a dual license. For non-commercial purposes, the terms of the LGPL 3.0 license hold. For commercial purposes, please contact us.

For our publication Detecting Similar Linked Datasets Using Topic Modelling we have the following additional material:

• For the first experiment, you can find the gold standard as well as the detailed F1 scores of Tapioca and a second version of Tapioca that uses the Jensen-Shannon divergence, in this folder.
• For the second experiment, you can find the detailed values of the P(w|T) and the A measure in this folder.
• For the third experiment, you can find the detailed values of the P(w|T) and the A measure as well as the F1 scores of our approach in this folder.