navigation: home || definitions || about


About the site...
Rhymes like dimes.

I initially made this site/app for the sole purpose of winning an argument. At a friend's suggestion, I decided to throw it online. For the most part, it's a mashup of various open source technologies, though I did end up writing the actual rhyme processing code myself... and that's really the meat of this sandwich, nahmsayin? (j/k). I wanted to use the fantastic Datamuse API (which powers the legendary Rhymezone), but doing so resulted in on the order of 300 API calls per verse. Since it's a REST API, these were all HTTP calls, making the app unacceptably slow. Ain't nobody got time for that. Instead, I wrote the rhyme processing code myself in PHP, using Martin van der Lee's phpSyllable class, Frank Liang's famous TeX algorithm, and Steven Woodbridge's Double Metaphone class. My code finds rhymes by dividing words into syllables, and dividing those further into an onset and rime (yes, I spelled that correctly), and by comparing metaphone keys (phonetic encodings). Apparently there's an even further improved version of the Metaphone algorithm called Metaphone 3, but it's not open source/free, and I wasn't tryna hear that. Lastly, lyrics are grabbed via the Genius API and those pretty graphics are thanks to PHPLot. I'm also using the Aylien API for certain calculations. Coming soon (maybe): ghostwriter identification/guessing, improved processing of multi-syllabic rhymes (yes, I know it's far from perfect right now), Billboard/top 40 charting info (if applicable), maybe a mobile app w/ Shazam integration.

About me...
Hollatchaboi.

I'm a mathemagician, computer scientist, freelance developer and tech consultant, doctoral grad student, startup veteran, vinyl collector, and sneaker fetishist, among other things. In a past life, I was a child actor; I also have a doppelganger from a parallel universe who's a dj and rapper. I went to school at upenn for my undergrad and master's degrees, and am now at uiuc tryna wrap up my phd.

I really don't see why my personal opinion matters, but since I've gotten a GRIP of messages asking, I really can't pick a single favorite rapper (or group or whatever). I listen to and appreciate a wide variety of hiphop ranging from indie backpacker rap to drill music, including everything in between. That said, if I had to pick a top 5 it would look something like (no particular order): Big Pun, MF Doom, Nas, Big L, Slug (Atmosphere), with honorable mentions to Qwel (Typical Cats) and Ghost & Rae (Wu Tang) as a duo.

I tend to stay away from most social media, but feel free to email me at rishi[at]inboxalias.com or check out my other website. Alternatively, you can use the contact form below.

Contact...
For those who are as lazy as me.

If you want to send me a quick email, fill out the form below. foxyform
  
 The Basics...
What up, son.

Songspin is a web app that calculates various lyrical statistics about user-specified hiphop songs. The stats calculated include average rhyme scheme length, average # of syllables per rhyming word, rhyme density, overall song sentiment/polarity (either positive, negative, or neutral, with an associated confidence level), and some other fun stuff like possible hashtags. It also displays graphs of the aforementioned stats, and is capable of both single song analysis as well as a comparison of 2 songs. To get started, click 'home' and enter the song(s) you wish to analyze on the left under the 'Getting Started' tab. If you'd like to incorporate Songspin data into your site/app, please don't try to hack together some screen scraping bullshit. I haven't made a public API or anything (yet), but just hit me up and I'll help you figure something out that doesn't strain my bandwidth and other server resources.