AI Bot App which Integration with RASA for Content Search

Overview

This project integrates an Android client with a RASA-based AI bot running on a server. The bot can respond to user queries and facilitate content searches using the information stored in specific directories.

---

Steps and Guidelines to Make It Work

Server Setup

1. Install Python:

   • Ensure Python 3.7, 3.8, 3.9, or 3.10 is installed. Recommended version: 3.9.13.
   • Download Python from the official site.

2. Install RASA Open Source:

   • Run the following command to install RASA using pip:

     pip install rasa

   • Verify the installation:

     rasa --version

3. Create a New RASA Project:

   • Run the following command to initialize a new RASA project:

     rasa init --no-prompt

   • Navigate to your RASA project directory.

4. Train the RASA Model:

   • To train your RASA model, use:

     rasa train

5. Test and Run the RASA Assistant:

   • Run the bot in the terminal using:

     rasa shell

   • Or, run the server with API support using:

     rasa run --enable-api

6. Training Data:

   • The model can be trained using files located within your project directory such as nlu.md, domain.yml, stories.md, etc.

7. Install spaCy for NLU:

   • Install the spaCy library using pip:

     pip install spacy

8. Install Other Required Packages:

   • Install any additional dependencies using pip as required by your project.

---

Client Setup (Android App)

1. Install the Application:

   • Download and install the Android application on the device.

2. Grant Permissions:

   • Ensure the app has the required permissions to access the device's storage.

3. Network Security Configuration:

   • The server system must be configured to allow cleartext traffic using the server's IP address. This is crucial for enabling communication between the Android app and the server over HTTP.

---

Guidelines

• Both the client (Android app) and server (RASA bot) should be able to communicate over the network.
• The bot can provide services regardless of how much it has been trained.
• All files needed for content search must be placed in a folder named Rasabot in your device's internal storage.

---

Working Process

1. User Interaction:

   • The user communicates with the AI bot (RASA) through the Android app for queries.

2. Content Search:

   • The bot provides instructions for content search, and the user's request is sent via an HTTP call to the server.
   • On the server side, the bot processes the request and sends the appropriate response back through the same communication path.

---

Tools Used

Client-Side

• Android Studio
• Java
• ITextpdf Library (for text extraction)

Server-Side

• Python
• RASA Framework
• spaCy Library (for NLU)

---

Contact

For further information, feel free to contact.