[PYTHON] I tried to make a "fucking big literary converter"
What I made for the time being
https://kusodeka.v-to-i.com
Try using it! !!
Overview
Needless to say, it is an application that makes you look like "Kusodeka Rashomon" when you enter text.
How are you doing
The code isn't publicly available, and the text conversion isn't very accurate, so I'll just give you a quick overview here. It's a very crude algorithm because I wanted to make it quickly for the time being.
- From the difference between Koshideka Rashomon and Rashomon, make a dictionary (Koshideka dictionary) of what kind of word is preceded by a word like "Koshideka" (let's call it Koshideka word) (Example: "Rashomon": [ "Big at the level of doubting sanity", "Fucking big", "Fucking big", "Tochi crazy fucking big"])
- Use Mecab for morphological analysis to identify the part of speech of a word in the input text.
- Insert a fucking deca word before a word with a specific part of speech, such as a general noun. At this time, select and insert the value of the key that has the closest similarity to the input word from the fucking deca dictionary (using gensim).
- Finally, evaluate the naturalness of the output text with the language model and check for any strange parts (not very well).
- The web app is made with flask.
Finally
I would like you to play around with it, but I would be very grateful if you could comment on things like "This will make things work better!" And "Why don't you add this kind of function?"
As for my impressions, I enjoyed being able to easily create my own web application, and I'm glad that I was able to put the NLP I was studying in the classroom into practice.