Semantive - Knowledge Mining with Cognitive Services

Demo: https://www.youtube.com/watch?v=BqSz9rU009Y

The Team 👪

1. Michał Chęciński
2. Bartosz Strachowski
3. Michał Kolendo
4. Maciej Wasik
5. Jakub Rejent

Use Case 🔍

Service to gather technical documentation for project or whole organization. The documentation includes:

• code
• scripts
• presentation
• project description, READMEs
• configuration files (eg. ARM templates)

In terms of the file types many filetypes can be used:

• Markdown
• json
• pdf
• Microsoft Office files (docx, pptx)

The solution

Functionalities of the system 🛰️

• Custom Skill - Bing search for entities
• Custom Skill - estimated reading time
• Custom Skill - template recognizing
• Display documents with immaculate preview of files of type:
  • png, jpg
  • markdown
  • json
• Searching big datasets of documents
  • filtering by parameters
  • paging
  • display document details
• Aggregate information regarding files
• Uploading new documents through the webpage
• Role Based Access to the system using Auth0
• Preview document on the page

Used Azure services 🖥️

The solution uses the following Azure Services:

• Azure Storage Account - to keep files with documentation.
• Azure Cognitive Search - to create search index on files.
• Azure WebApp - to access search and upload functionality.
• Azure Functions - to host custom skills for the Azure Search Indexer.
• Azure Cognitive Services - to extract additional data for text and images, also another Cognitive Service is used to get data from Bing WebSearch for one of the custom skills.

Cloud architecture 🗺️

Technical stack 🛠️

• Microsoft Azure
• ASP.NET Core MVC
• jQuery + Bootstrap

Reconstructing the solution 👷

To reconstruct the solution you can follow the steps described in the How to reconstruct the solution document.