Hausaufgabe 4 Abgabe bis 8.1.17 1- Compare the performance of the models we have implemented so far: - Character-based POS model - Word-based POS Model without per-trained embeddings - Word-based POS Model with per-trained embeddings - Joint POS Model (Char+Word model ) without per-trained embeddings - Joint POS Model (Char+Word model ) with per-trained embeddings What is the best model ? Is it possible to boost the performance even further? Chunking: Text chunking consists of dividing a text in syntactically correlated parts of words. For example, the sentence He reckons the current account deficit will narrow to only # 1.8 billion in September . can be divided as follows: [NP He ] [VP reckons ] [NP the current account deficit ] [VP will narrow ] [PP to ] [NP only # 1.8 billion ] [PP in ] [NP September ] . 2 - Build a neural chunker based on the data we provide for chunking.