added encoding = "utf-8" for csv reader for the Link Prediction and Reconstruction file reading
modified find_matching_terms() to bring best possibilities in the model's vocabulary closed to the input word, instead of considering only words starting by the input words, leading to "KeyError" error
added encoding = "utf-8" for csv reader for the Link Prediction and Reconstruction file reading modified find_matching_terms() to bring best possibilities in the model's vocabulary closed to the input word, instead of considering only words starting by the input words, leading to "KeyError" error