Hack-Jaipur

VidAlytics - Video Analysis for Marketing and Business Development

It is an online service created by team Digital Chaos for content creators who perform video content analysis for designing effective marketing and business development related policies. It supports various local languages too, including Hindi. Also, a chrome extension which helps you navigate to the part of the video having the content you are looking for by just typing the keyword

1) WebApp

Screenshots

About

• Project Name: Vidalytics
• Short Project Description: Video Analysis for Marketing and Business Development
• Team Name: Digital Chaos
• Team Members:
  • Jay Mehta https://github.com/jaypajji4444/
  • Deep Gandhi https://github.com/deep1401/
  • Jash Mehta https://github.com/jashmehta3300/
  • Nirali Parekh https://github.com/nirali25parekh
• Demo Link:https://youtu.be/_QtbxXm0fkQ
• Labels: Machine Learning, Azure, Praw, Video Analytics, Node.js, Flask

Installation

git clone "https://github.com/jaypajji4444/Hack-Jaipur"
pip install -r requirements.txt
cd codebreak
python app.py
cd ..
cd reddit
python app.py
cd ..

## Starting app
node server.js
open: localhost:8000

Description

Content creators over the country suffer from a common problem that cannot be solved by being creative. In today's cut-throat marketing world, it is absolutely necessary to ensure that your product is marketed to the right audience in the right way. Video Indexer consolidates various audio and video artificial intelligence (AI) technologies offered into one integrated service, making development simpler. By abstracting away all the dirty API calls and loading screens, we aim to create a clean, one-stop service that would be of immediate help to anybody looking to market their product. To achieve this, we provide three main services - Label Identification (Video analysis), Keyword Identification (Audio analysis) and Sentiment Analysis. By combining insights from these channels, we provide a list of Reddit communities which the user can then target for marketing their content effectively. Also, by identifying the sentiments within the video, the user now gains an upper hand by getting exact target markets on where to spread their content. Also, the smart sentiment filtering helps users control the tone of the message being sent across. This is not only limited to the YouTube content creators but has widespread applications for various other entities such as companies looking forward to launching their products, people trying to convey a message be it in an interview or social. We look forward to integrate Facebook Graph in the app by finding a workaround to the privacy policies. Thus, we have created a one-stop shop for video analytics.

Technology Stack Used

• Node.js
• MongoDB
• Flask
• Azure
• Bootstrap

Note

We used the Reddit Praw data mining tool for the dynamic mining of data every time a keyword is generated. It helps keep up with the current sentiment of society.

2) Youtube_Extension

A chrome extension which helps you navigate to the part of video having the content you are looking for.

Use Case

1. Find a relevant video without any need to watch it
2. Helps in marketing by analyzing your video and including relevant keywords
3. Saves your precious time

Getting started

Backend(API)

pip install -r requirements.txt
python app.py

Alternatively, Run the build.sh file as :-

bash build.sh

Running Extension:

1. Go to chrome://extensions
2. Make the developer mode on(top right corner)
3. Select load unpack package (top left corner)
4. Browse and select the client folder of the project
5. Boom you will see the app in extensions bar...

Screenshots

On opening the extension

When you search for something

On clicking the result

_