Poetries has been one of those finest arts which make you feel the emotions at very first glance. There have been many great Poets in our history like whether it is William Shakespeare or Emily Dickinson, they create beauties with the immense hard work.
But it’s sad to say, the world will no longer need their hard work or we can say production of art might in future be more quick, unique and cheap because of the tremendous growth of the fields like Natural Language Processing and AI, and they are still growing.
Computer scientists the University of Melbourne in Australia and the University of Toronto in Canada have developed an algorithm that is capable of writing poetry following the rules of rhyme and meter.
With the use of poetries rules and taking the meter into account, this AI algorithm creates weaves of words and grouped them together to produce meaningful sentences.
This AI is trained extensively on the rules it needed to follow to craft an acceptable poem and the dataset researcher used to train the AI has over 2,600 real sonnets.
The algorithm analyzes them to teach itself how the words worked with each other. Within no time it soon starts to mimics the iambic pentameter and rhyming pattern of the poems most famously written by ol’ Bill Shakespeare himself.
The team uses a combination of convolutional and recurrent networks for modeling Chinese poetry, and also use an incorporating an attention mechanism and training at the character level.
For English poetry, the team introduced a finite-state acceptor to explicitly model rhythm in conjunction with a recurrent neural language model for a generation.
Hopkins and Kiela, the team members improve rhythm modeling with a cascade of weighted state transducers and demonstrate the use of character-level language model for English poetry.
The researcher further jointly models both poetry content and forms, and unlike previous work which uses dictionaries for rhyme, so that the AI can learn how to generate them automatically.
The neural architecture of which AI used is, composed of 3 components: (1) a language model; (2) a pentameter model for capturing iambic pentameter and (3) a rhyme model for learning rhyming words.
If we talk about the results, then the rhymes and meter in the machine-generated poetry were more precise than in human-written poems. It gets successful in tricking many humans into thinking that its poems were penned by AI.
Further to test its Poetry writing capabilities the researcher puts the AI into a battlefield against humans. The bot’s verses were mixed with human-written poems and then scoured by volunteers, the results were surprising the readers were split 50-50 over who wrote them.
We hope that this which will definitely boost the confidence of this AI. Some improvement is still needed related to errors in wording and grammar. Researchers want to fine-tune the algorithm so that it sticks to a single topic, or design an algorithm that can write short fiction.
Jey Han, one of the researchers on the project said: “While the application itself may not seem directly relevant to real-world applications, the underlying machinery of our model shares the same core algorithm that drives other problems that require generation.”
Adam Hammond, a University of Toronto English professor said “I’m excited to see what’s possible, but I’m very skeptical. I think it’s quite easy to have a deep learning model to spit out lines of verse in rhyming iambic pentameter. It’s much, much, much, much harder to train it to have an opinion, or a feeling, or a desire, or a story to tell.”
If you are interested in reading the paper, it’s here.
More in AI :