JowsDatalog

Jointly-Weakly-Sticky (JWS) Datalog+/- is an expressive member of Datalog+/- programs. It is characterized by marking procedure, the existential dependency graph, and joint acyclicity. Query-answering (QA) can be done in polynomialtime in data through SChQAS, a chase-based, bottom-up QA algorithm for JWS Datalog+/- programs. The QA algorithm can be optimized by using a magic- sets query rewriting technique, MagicD+, for JWS programs. MagicD+ takes a Datalog+/- program and a query, and rewrites it into a new Datalog+/- program. The new program can be the input for SChQAS. With the new program, SChQAS avoids generating irrelevant facts. Designing an Ontology-based data ac- cess (OBDA) with which QA can be done in polynomial-time in data complexity for JWS Datalog+/- programs is an open area of research. The main objective of this thesis is the design and implementation of an OBDA tool, JowsDatalog, in which SChQAS and MagicD+ are implemented.

Prerequisites

The user must provide an database connection information, including username, password, db address, db port, and db name. The recomended RDBMS software is PostgreSQL. JowsDatalog is a java based tool and needs JRE.

How to run the binary

Executing JowsDatalog can be done by running following command from command line:

$ java -jar /PATH_TO_JOWSDATALOG_DIR/JowsDatalog.jar

But, before running this command make sure that you have edited the config file.

Config File

This file has following parameters which must be provided by the user:

• db_username: The database username (i.e. postgres)
• db_port: The database port (i.e 5432)
• db_password: The database password
• db_address: The database address
• db_name: The databse name
• table_prefix: JowsDatalog creates table for each predicate. We can add a prefix to each table.
• attribute_prefix: JowsDatalog creates table for each predicate. We can add a prefix to each attribute of a table.
• resumption: Resumption is one part of SCHQAS. The value should be either true/false
• number_resumption: the number of resumption (an integer)
• reset_db: Either true/false and indicates that the tool build the database or not. This help us to use the materialized instance after.
• repair: future feature, set false.
• default_weight: future feature, set 1.
• program_dir: This can be a directory or a file which points to a Datalog+/- program(+).
• query_dir: This can be a directory or a file which points to the queries defined on top of the program.
• csv_dir: The directory where CSV files are stored.
• schema_dir: The direcoty of the schema of CSV files.
• use_magicSet: Whether the tool should use MagicD+ or not (true/false)
• use_basicApproach: Whether the tool should use basic approach or not (true/false); preferably false
• load_CSV: Whether to load CSVs or not (true/false).
• printQans: Whether to show the answers to queries or not (true/false).