Octopost

Octopost is a web application, where everybody can freely write posts and vote for posts others have already written. Those posts can be sorted with a variety of different filters. What makes Octopost special is, that every post will be given a class classified by a RNN implemented with TensorFlow. The technologies used in this project are ASP.NET Core, Flask, TensorFlow and Angular.

How to run?

Train the RNN

Execute the following command in Octopost.TopicClassifier:

python text_classification.py

Run the RNN

Execute the following command in Octopost.TopicClassifier:

python text_classification.py --server run

Run the Web Api

Configure ports, connection string and location api key in appsettings.json

If UClassify Api Key is not set, local Python server will be used for classification Set the environment either to 'Development' or 'Production':

set ASPNETCORE_ENVIRONMENT=Development|Production Start the Web application on port 5000

Run the application from the dotnet CLI:

dotnet Octopost.WebApi.dll

Run the Web UI

Execute the following command in Octopost.WebUI:

ng serve

Run Mobile App (unfinished)

Execute the following command in Octopost.MobileApp:

ionic serve

Build the Web UI

ng build --prod --aot --verbose --build-optimizer