How to find syntactic dependencies in text using unsupervised method and context information?

I know there are ready libraries to find syntactic dependencies and besides supervised methods, I have studied some of the unsupervised dependency parsing which uses POS tags and other mathematical and statistical techniques to solve the problem.

I am working on a challenge to find out that is there any way to find syntactic dependencies in an unsupervised way and only by using the co-occurrence of words with each other and their context information? For example is there any way to use word co-occurrence and make a graph of words which then by using Clustering methods we group them and show the syntactic dependencies? all idea about finding syntactic dependencies which is based on unsupervised methods and using graphs appreciated.