Inspiration
We thought it would be fun, most online phonetic converters that we found online were good, they did not convert to phonetics that the average person could understand. We were also not able to find any that allowed the user to input a URL for a website they wanted to convert and we wanted to create that.
What it does
Our program takes either a text input from the user, or extracts the text from a user-given URL, and converts it into phonetic text which is then outputted to the user.
How we built it
We found a dictionary of words with their ARPABET spellings from Carnegie Mellon University(1). We used python to turn this list into an indexable structure of terms and ARPABET spellings. We then used the Haskin's key (2) to find the IPA values corresponding to the ARABET and then used Wikipedia (3) to find the AHD values for the IPA. Finally, we compiled another .txt file of ARPABET to AHD and used the same python function we'd built for the first dictionary to turn it into a function. We then built a function that takes text and uses the two dictionaries to convert it to the AHD spelling. Finally, we built a function to take URLs and extract the text found on them and make it readable for our converting function so that a user can enter a URL and get the entire site back spelled phonetically. For this, we used an algorithm for using Beautiful Soup (4) and then expanded on it by cutting down on excess extracted text.
- http://www.speech.cs.cmu.edu/cgi-bin/cmudict
- http://www.haskins.yale.edu/staff/gafos_downloads/Arpabet-to-IPA.pdf
- https://en.wikipedia.org/wiki/Pronunciation_respelling_for_English
- https://matix.io/extract-text-from-webpage-using-beautifulsoup-and-python/
Challenges we ran into
There were two main challenges we ran into. As stated above, we had to convert a list of words and their phonetic spelling into a dictionary, this was challenging because neither of us had much experience with python and it turns out it is complicated operation to perform. Another challenge we faced was extracting the text from websites based on their URL. As stated above we ended up using an algorithm made by Beautiful Soup and editing it to fit our needs.
Accomplishments that we're proud of
We're proud that we used python even though neither of us is experienced with it and we're proud that it can take a URL as input because most phonetic lists only accept raw text. Additionally, we think our English friends would be impressed that we did a project involving words.
What we learned
Throughout this process, we learned a lot about Python, which was a new language for us. We also learned how to access data from a URL and then parse it to useable text.
What's next for fŭnĕtĭk kŭnvərtər (Phonetic Converter)
We hope to convert this to a website and improve the URL input capabilities to accept more complicated sites.
Log in or sign up for Devpost to join the conversation.