Semantic Table Linker

Semantic Table Linker (STL): A web-based application (a server part) to annotate tabular data.

Version

1.0

Preliminaries

Source tabular data are vertically oriented tables in which the data are systematically aligned in columnar sequences. Each sequence may incorporate a descriptive label (header) to clarify its content.

Assumption 1. Table headings must be single-level and placed in the first row of the table.

Assumption 2. Source tables may not be normalized (non-relational tables).

Installation

First, you need to clone the project into your directory:

git clone https://github.com/YRL-AIDA/semantic-table-linker.git

Next, you need to install all requirements for this project:

pip install -r requirements.txt

We recommend you to use Python 3.7 or more.

Directory Structure

• examples contains three sets with example tables (pubmed-cancer, wikipedia-bird, wikipedia-universities-in-the-kingdom);
• stl contains main scripts for annotating tables based on the DBpedia knowledge graph.

Usage

Console mode

In order to use the STL in console mode, you may run the following command:

python main.py

Run this script to process source tables in CSV format. Tables must be located in the datasets/test/ directory.

The processing result are presented as JSON format and will be saved to the results directory (json and provenance subdirectories).

Web mode

In order to use the STL in web mode, you may run the following command:

python app.py

NOTE: This mode requires a client part of application (stl-client)!

Author

• Nikita O. Dorodnykh