implementing Word Embeddings for adjectives, obtained embeddings to preserve as much synonym relationship as possible.
We provide only the raw form of data to be used in this project. For embeddings, you are required to process this dataset using spaCy. You can write out your processed data file in the Present Working Directory (PWD), which can later be used to train the embeddings.
While the vanilla word2vec skip-gram model generates its (mini-batches of) training data between a center word and a random sample of its context, you are encouraged to think of new methods such that the resulting embedding for adjectives are more likely to preserve the synonym relationship in its top-k results.
4 freelancere byder i gennemsnit $186 på dette job
Having experience of python, I can do what you want. Relevant Skills and Experience python, software architecture Proposed Milestones $200 AUD - 3 days
I've already developed several w2v models in tensorflow.. Moreover I have sound knowledge in NLP too, so if you require anything else...it can done as well! *(bid provided is per task)