Igor Furgała
Michał Przyszczypkowski
Daniel Staśczak
Krzysztof Stronka
How to keep track of different problems people have in the organization?
How to optimise finding already solved similar problems?
We present Omnibot for Slack!
How it works?
Step 1: Collect meaningful technical terms
We collected 5000 mostly used tags from StackOverflow using StackExchange Data Explorer
Step 2: Describe each question by n-dimentional terms vector
Tokenized & stemmed using NLTK - Natural Language Toolkit to find technical terms
Represent terms found in the question using n-dimentional vector
Step 3: Show previously answered questions represented by vectors with highest cosine similarity