We are doing a project in my pattern recognition class. We want to train a neural network to learn how to recognize the unique style of a writer.
We plan to use some basic features such as word length, frequency, sentence length, paragraph length and variances of these variables as input data (We are aware of the naivety of our approach, but there is no linguist in our class). I looked at the Grok API and had a feeling that there are classes in it that can help us to prepare the input data. Please confirm. We also plan to use the number of clauses and infinitives to represent a level of sophistication in style. Can Grok help us to calculate this? We have a Java programmer in our bunch and he can help to do the math if Grok can detect words, sentences, and sub-sentence structures.
Thanks
JD
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Grok can detect words and sentences, but doesn't have any sub-structure identifiers at the moment. I would suggest looking at http://gate.ac.uk, as it may be more suitable for your task at present and we may be switching over to using the Gate architecture for Grok components anyway.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
We are doing a project in my pattern recognition class. We want to train a neural network to learn how to recognize the unique style of a writer.
We plan to use some basic features such as word length, frequency, sentence length, paragraph length and variances of these variables as input data (We are aware of the naivety of our approach, but there is no linguist in our class). I looked at the Grok API and had a feeling that there are classes in it that can help us to prepare the input data. Please confirm. We also plan to use the number of clauses and infinitives to represent a level of sophistication in style. Can Grok help us to calculate this? We have a Java programmer in our bunch and he can help to do the math if Grok can detect words, sentences, and sub-sentence structures.
Thanks
JD
Grok can detect words and sentences, but doesn't have any sub-structure identifiers at the moment. I would suggest looking at http://gate.ac.uk, as it may be more suitable for your task at present and we may be switching over to using the Gate architecture for Grok components anyway.