MASTER THESIS WEB APPLICATION

This project follows a SaaS architecture for providing analytical services to two different stakeholders. Offers to restaurant owners a tool for getting an analysis of the social opinion of their business. And provides to citizens a tool for letting them know about the restaurant ratings on their surroundings.

For achieving this, social networks data, an other available Open Data sources on the Web are queried. Multi-language user opinions and other unstructured data about the products and services of a restaurant business are obtained and processed through different semantic analysis techniques. The data is stored as a new graph model which is adapted to the channel of the stakeholder.

Required applications

• Neo4j
• Python 3.x
• Python VirtualEnv

Setup

1. Create a virtual environment for installing and executing the application:

$ virtualenv -p /usr/local/bin/python3 venv

2. Create a load_config.sh file from the template load_config_sh.dist and fill the variables with your development credentials.

3 Load the virtual environment and set variables:

$ source venv/bin/activate
$ source bin/load_config.sh

4. Install the libraries from the requirements.txt file:

$ pip3 install -r requirements.txt

5. Execute main.py file.

Deploy

All variables of the bin/load_config.sh.dist file have to be set up in the server. Furthermore, Textblob corpora has to be downloaded, as the bin/install_textblob_corpora file suggests.

$ git push heroku master

Integration Tests

Before executing the tests, some information will be loaded into the database. So, do not launch the tests on a production server because it could lead to data loss.

## Stop database
$ brew services stop neo4j

## Prepare environment
$ rm -r $NEO4J_HOME/libexec/data/databases/<DB_NAME>
$ neo4j-admin load --from=test/test_database.backup --database=<DB_NAME>

## Start database
$ brew services start neo4j

## Launch tests
$ python tests.py -v

The Integration Tests can be executed along with a code coverage test. The following commands will execute the integration tests. After that, a report with the code coverage of the application will be shown.

$ coverage run tests.py -v
$ coverage html

About

Jagoba P. G. jagobaperez92@opendeusto.es | https://jagobapg.eu