Skip to content

Latest commit

 

History

History
28 lines (15 loc) · 913 Bytes

README.md

File metadata and controls

28 lines (15 loc) · 913 Bytes

Understanding the Role of Background Knowledge in Predictions; Ecotoxicological Effect Prediction. Master's thesis by Nils Petter Opsahl Skrindebakke.

Install

conda create --name myenv --file spec-file.txt

Directory Structure

TCE: Triple Context Embedding in Python (under development).

data: Contains the training and test data.

graph_analysis: Tools for analysis of the KG.

kg: Contains the original KGs.

kg/processed: The results from the crawls (ie. scored/filered KGs).

obj: Contains a binary file used for cache when mapping to CID.

plots: Tools for plotting graphs and printing the results.

prep: Tools for downloading and and processing KGs, test and train data.

results: Binary file containing a dictionary with the results from the runs.

src: Contains the ML-models and other related tools to run and evaluate the predictions.

src/crawl: Contains the crawling and scoring algorithms.