Final Project Writeup

My Final Project can be found here (need to be on a Brown University google account to view):
 https://docs.google.com/document/d/1IOAkrvK49gDC2O34E18l8m_ybd6oVrtcVeo6L022O6U/edit?usp=sharing

The code repository can be found here:
https://github.com/jramos70/DLA-Project

The final project can be found here:
https://jramos70.github.io/final_project/intro.html

Project Update (pt 2)

A link to my progress is here: https://jramos70.github.io/final_project/intro

What I have done so far is starting to create the html files for the text game. In it I have written different scenarios that tell the story. I also experimented with adding interactivity to it. For example, one scenario has 3 different choices and only 1 lead to the next scenario (the other 2 lead to a game over screen). I used javascript to randomly choose which one is successful, so it will be different each time. I also added a scenario where you are in a space battle with another ship and you need to fire your "weapons" (in this case, clicking a button") at the other ship to progress to the next round.

I think my progress is a good example of what the game will look like. My next steps will be to add a lot more scenarios to the game and add some css for visual aesthetics. However, I do not want the game to look too polished, because I really want to make it feel more personal and hand-crafted, like the hand-written web article we read in class. Additionally, I played around with making random text from a dataset, but I didn't really like the results, so I decided to write the story myself.

Project Update

For my project update, I have started to put some initial thought into how my game will be structured. I have decided that for the decisions players have to make,  the outcome will not always be the same in every game. For example, if a player is given choice A and choice B, choice A will give different outcomes each time.

Another thing I am planning to add is items or companions for the player. Once again, these things will be randomized throughout the game so players will receive different items for each run.

I think the randomization of the project makes it more interesting as it isn't always the same game throughout each run. Additionally, I hope that by adding items to the game, it adds a layer of strategy to it.

Week 10 Experiment

My Experiment was to translate "Through the Park" by Nick Monfort (found here: http://nickm.com/poems/through_the_park.html)

Below are the results:

Durch den Park
Das Mädchen grinst und schnappt sich einen Müsliriegel. ... Das Mädchen zieht sich ein schlampiges Kleid an. ... Eine Wolfspfeife ertönt. ... Eine Wildblume nickt, fest umklammert. ... ein Kichern weht durch die Luft. ... Der Mann stürzt ab und hinterlässt den Vorwand. ... Tauben streuen. ... das Grün des Parks ist grau. ...

I chose German because it was a language I am somewhat familiar with. Overall, the translation ended up being fairly accurate and kept a lot of the original meaning behind it.

I then decided to translate it into Tagalog:

Sa pamamagitan ng parke
Ang batang babae ay nanunuya at nakakuha ng granola bar. ... Ang batang babae ay naglalagay sa isang slutty na damit. ... Isang lobo sumisipol tunog. ... Isang wildflower nods, mahigpit gripped. ... Isang giggle weaves sa pamamagitan ng hangin. ... Ang lalaki ay nagtagas, na iniiwan ang pagkukunwari. ... Pigeons scatter. ... Ang berdeng parke ay kulay-abo. ...

In this translation, some of meaning gets lost/convoluted. From this experiment, I can see that perhaps certain languages translate better into others, which makes sense given the origins of English and German compared to English and Tagalog.

Final Project Proposal

Project Title (Temporary): Space Adventure

Summary: Space Adventure is an interactive web story that allows users to walk through a science fiction story online. At each prompt, users will be given a set of choices. Each choice they make will determine how the story progresses for them. Because of this, no two storylines will be exactly the same, which will lead to a more interactive experience!

Keywords: Interactive, randomized

Would I like to present: Unsure at the moment

Details:
Concept: The concept is somewhat of a continuation of my Week 8 Experiment. I find story driven games to be very fascinating and I think making them interactive adds an additional layer of immersion. I believe combining language and technology makes the storyline more exciting and allows for the story to be non-linear.

Implementation: I primarily will be using HTML/CSS/Javscript to create the game. I am also looking into the possibility of using a GAN or Markov Models to generate the story prompts so that no two games will be the same.

Inspiration: My inspiration is text games I have played in the past, such as Zork: https://www.youtube.com/watch?v=PWQDccL0aXM

Project Plan: My plan is to create a simple web app using HTML/CSS/Javascript to display the game and add visuals and interactive buttons. The main challenge will be to dynamically generate story prompts (if I decide to go that route).

Week 9 Readings

I thought that the "Machine Translation and Global English" paper brought up really good points about the state of NLP projects and the problems we are trying to solve using NLP. I like the idea of having a "Global English" language because it solves some of the big problems in translation, mainly, meanings get lost through translation. I thought that this was interesting because in my week 1 experiment, the meaning of the lyrics got lost after I translated it multiple times through Google Translate.

Similarly, I thought "Algorithmic Adaptations" was pretty cool because it was similar to my project as well. It explores how the translation algorithm translates things by utilizing blanks. I thought it was a pretty cool idea because I didn't realize before how missing words can drastically change the meaning of a sentence and that by excluding them, it can be tricky for the translation algorithm to get a proper read on the meaning of the sentence.

Week 8 Reading Response

I thought that the "Handmade Web" article was very interesting because the culture of the web today is much different than it was 20 years ago. In the past, the internet was a lot less commercialized and there wasn't many frameworks/libraries available to create websites. Many of the websites were crafted using basic HTML/CSS/Javascript. In the present, the web is so commercialized with ads on every site. Additionally, there are so many web frameworks available that standardizes the way the web is programmed, which is some ways can hinder creativity.

Reading "Aesthetic Animism" there is a thought that perhaps AI could replace human creativity. However, I believe that human thought is special and cannot be 100% replicated. Most of machine learning is learning from past human examples and really only represents what the AI thinks humans think rather than what the AI actually thinks. The conversation of whether AI is thinking for itself is an interesting one, but I believe that because poetry requires human thoughts and emotions, AI could never truly replace human poetry.

Week 8 Experiment

My experiment for this week is a text based game that I created. My inspiration for it was some text based games I used to play as a kid where choosing different outcomes leads to a different story.

Experiment found here: https://repl.it/repls/UglyTurbulentCoin

Week 7

I believe that the key difference between movies and theater is that the audience can control the movie as it is pre-recorded, but theater is a live production. Whether or not technology is included does not change this description.

Although some people believe that technology turns theater into a sort of psuedo-cinema, where the quality of the performance isn't as good as either a traditional theater performance or cinematic film, I believe that technology can be used to enhance the production and make a more interesting live performance. One example of this is  the use of projectors can be very useful because it can produce moving images while the actors in the play perform. Overall, I believe that the use of digital arts can enhance the production value of any medium and add a layer of interactivity to make a more immersive experience for the viewer.

Week 7 Experiment

For my experiment, I decided to look at gender diversity in the CS Department. For my dataset, I took the text of all of the students who are TA's this semester and classified them as either male or female based on their names. Below are the results.

Figure 1. # of TA's by gender for each course level

Figure 2. # of HTA's by gender for each course level

Figure 3. # of HTA's by CS Field (Upper Level courses only)

In figure 1, we can see that for intro level courses, there are an fairly even number of male and female TA's. However, in intermediate and upper level courses, there are almost double the number of male TA's compared to female TA's. In non-concentrator courses (cs0020, cs0030, cs0100) there are slightly more female TA's than male TA's.

I then wanted to take a look at gender diversity for the head TA's. There are significantly more male TA's and female TA's for intro and upper level courses, but the male-female ratio for intermediate and non-concentrator courses are fairly even (3:2 and 2:4 respectively).

Since the differences in male TA's to female TA's was so high, I decided to look at the different fields of CS to learn more about why this may be the case. I discovered that there were significantly more men than women in AI/ML courses (43 men to 12 women). There are also almost double the number of men in theory courses. Thus, ML/AI and theory courses at Brown seem to more more male dominated compared to other field in Computer Science.

I thought this project was really interesting and sheds some light on how the Brown CS Department is doing in terms of gender diversity. For a future project, it would also be really interesting to plot similar data, taking race into account.

Week 6

Experiment

My experiment can be found here: DLA Experiment - Week 6
My experiment is a blend of language and code related to Halloween and Spooktober!
Readings

I thought "Opinions of my Own" was a really cool Chrome Extension and a great idea. I feel that in today's society, many companies and other entities are harvesting our data for their own purposes. I believe that, as a society, we do not fully appreciate the importance of our digital data/privacy, which allows third parties to gather our information and invade our personal privacy. I hope that more people will be aware of how important their data is and take more measures to protect it.

I think that language becomes data when the original meaning/intent of the language becomes stripped away. For example, a person's twitter account gives them a means of expression. However, if someone is scraping their tweet history, each tweet is just another data point rather than a form of language.

I believe we can ethically use online textual data by only using public sources where the user(s) agree that their data may be used. In these cases, the user should be fully aware of how their data is used and the person using their data must explicitly tell the user(s) what data they are using and how they are using it.

Week 5

Experiment

My experiment can be found here: https://jramos70.github.io/week5/experiment.html. My project is called Elon-spirational Quotes and it displays various Elon Musk quotes in 3D. The three keywords I thought of while making this project is change, non-static, and pop out.

Readings

I thought that Course in General Linguistics was an interesting read as it says that speech and sound alone do not have any meaning; it is the shared experience(s) between two individuals that allow those things to have meaning. I think that makes sense as someone who speaks English cannot effectively communicate with someone who only speaks Spanish because they do not have the shared experience of speaking the other's language.

Related to the above idea of what defines meaning in a language, Mez Breeze fuses "English, poetic conventions, programming code, contemporary social commentary, and online communiqué". Although parts of the language can seem foreign to certain viewers, people with a background in any of those subsets of the language will be able to understand those portions of the text. It is there past experience(s) that allow them to interpret the meaning of Mezangelle.

Week 4

Experiment

My experiment can be found at this link: https://jramos70.github.io/week4/experiment.html

My experiment was to take 4000 Elon Musk tweets and 4000 Donald Trump tweets and use Markov chains to mix their style of tweets. I chose Elon and Trump because they are two very active twitter users who usually tweet about very different topics, so I thought it would be interesting to see how their styles would mix.

Readings
TextRain seems like a really cool use of Augmented Reality and it is cool how the words dropping actually form a poem rather than random letters. Stream of Consciousness had a similar idea, except the words move in a stream rather than fall down vertically. Additionally, the artist can control how the stream moves. It is cool to see how two similar ideas get modified to make each art piece unique.

I would like to try out Curved out of these three art pieces the most because it seems very engaging. The mix of audio and visuals would make the experience really immersive.

For me, these 10 keywords make 3D text different from 2D text: Immersive, engaging, active, moving, change, different, physical, non-static, realistic, pop-out. When approaching a piece of language art in 3D, I would see try to visualize how the piece would be in 2D and think about how different the experience would be. Additionally, I would also try and think about "why" the artist decided to use 3D instead of 2D (what are the benefits or pros/cons). I also want to think about how my other senses are affected by the 3D piece (like in Curved where the audio changes the experience).

Week 3

Experiment
My experiment for this week can be found at this link: Experiment. My experiment name is "Fun Balls" and it creates a new ball every time the user clicks inside the canvas. My inspiration for this project was screen savers on computers.

Readings

I found the bot or not website to be quite fun and found that it was hard to determine whether or not a poem was written by a bot or person. I feel that since the bots are programmed by humans, in a way, the poems are still "human-written". This could be potentially why it is hard to distinguish human and bot written programs.

I thought the context free grammar was interesting because it formalizes a way to create sentence structures. It also makes me wonder if the bots used in "bots or not" uses this structure or if they use a grammar-free structure. Additionally, if personal home assistants like Alexa use these language rules, does it make it easier or harder for them to interpret improper grammar or slang from users?

In "Digital Oulipo", it was interesting to read how early practitioners of digital language arts used math to form literature and poems. Coming from a quantitative background, I always thought of math and language arts to be quite separate as math relies on logic whereas art is more free-form and emotional. Thus, it is pretty cool to see how art is created through a mathematical process.

Week 2

Experiment:

For my experiment, I took the lyrics from the song "Killshot" by Eminem and ran i through Google Translate multiple times and collected the results. The order in which I translated the lyrics was:
English -> Swedish -> Hebrew -> Vietnamese -> german -> Thai -> Icelandic -> Spanish -> Japanese -> Norwegian ->English. I then used text to speech to convert the lyrics into audio. Afterwards, I played the text to speech audio over the song's instrumental.

It's interesting to see how different the lyrics of my new version is from the original. An interesting thing to note is that "Killshot" is a diss track, where Eminem berates and mocks his rival (MGK). However, in my version, there is a lack of vulgarity and, in some cases, it almost seems as if Eminem is being quite friendly towards MGK. The dramatic change of lyrics between the two versions highlights how imperfect Google Translate is.

Link to original song: https://www.youtube.com/watch?v=FxQTY-W6GIo
Link to my experimental version: Killshot by Eminem, translated 10 times

Readings:

I found the LOLCat and Doge reading to be quite humorous. I never thought I would see a linguistic essay about memes. It is interesting that LOLCat emerged from funny pictures of the animal in general whereas Doge came from a specific dog picture. I think its interesting that the author calls Doge the "antithesis" of LoLCat because I believe the frequent creators of these memes most likely enjoy both. I think that both memes are humorous in their own ways and that the incorrect grammar and spelling adds to their charm.

FlatLogic talked about the shallowness of interacting with technology, but I feel that advances in technology are making tech interactions feel more lifelike. With all the research going into virtual reality, augmented reality, 3D printing, etc. I feel that technology is becoming more and more immersive. Additionally, with the physical immersion, I feel that these technologies will also add emotional depth because the user will feel more connected to the experience.


Art:
https://www.instagram.com/p/BmcrF6RFxcI/?taken-by=internet_poetry. This particular piece is interesting because it lacks punctuation.

http://vispo.com/uribe/metalico2.html - I actually found this one amusing because it uses letters to emulate a zipper opening and closing.

Week 1

"Principles of New Media"  talked about continuous an discrete media. For example, a song (in real life) is continuous because there are no breaks, but in a digital photo, everything is made up of pixels and everything can be represented in a numerical form. Because new media is programmable, we can sample continuous intervals and make it discrete. Modern digital media technologies are discrete because language is discrete; Sentences, words, and morphemes are all discrete.

Additionally, modern media technology is both modular and discrete. That means, that an artist can combine singular objects together to create art. In terms of variability, it makes it easy for artists to split up other works in to form new pieces and make a unique piece of media themselves.

In"Prosthesis", Ian has a collection of songs that sound very choppy and robotic like. A lot of the songs sound like he's doing programming syntax (e.g. for loops, i++, recursion, do while) and he builds upon those elements. In this way, he uses discrete objects to build something greater.

"The Listener" is an alexa skill that performs a conversational performance. It is interesting because alexa has a voice that seems human, making it seem like the user is talking directly to the artist. However, Alexa is purely synthetic and is an example of human aurality. The human-like features of Alexa makes it invited to the household and almost makes it part of the family.

In "Long Rong Song", OTTARAS repeats "LONG RONG SONG TONG", except uses different vowels each time. He uses discrete elements, but in a way "programs" it to be slightly different. Additionally, he uses discrete objects (in this case, words) to create a song.