Week 9: Digital Karnak

During my freshman year I took Architecture 10A with Diane Favro. The course covered a lot of Greek and Roman architecture. I also distinctly remember pulling an all-nighter to complete an interactive Google Earth guide through Roman victory marches. The project took the viewer through the city of Rome on Google Earth past different landmarks. Movement though space is one of the most important ways we can understand architecture so to have a way to digitally interact with a space is very helpful. I think that Favro’s “Meaning in Motion” can be applied to another UCLA project is Digital Karnak (Favro is one of the directors of this project). Digital Karnak is a great exploration of the temple complex at Karnak. One of the things that makes this project so great and really distinguishes it from other 3D interactive projects of ancient spaces is that you can explore what the complex looked like through time. The timemap indicates what buildings were built, destroyed, or inactive during a certain period of time.

The experience Karnak tab includes information on how the space was used for different activities. Unfortunately the way we explore these spaces is only through 3D videos that don’t allow the viewer to freely explore the space. However, this isn’t a problem as these are supposed to be like “guided tours” through a specific place during a specific event. Fortunately this is supplemented by the Google Earth tab if you want to download the 3D model on Google Earth so you can freely explore the complex. Unlike the Digital Roman Forum The information is very easy to navigate and you don’t have to search for a specific thing in order to find information on it. There isn’t a search bar in Digital Karnak but you can use the “Browse archive” tab. I think that Digital Karnak exhibits the five operating principles set forth by the CVRLab which are to “follow the highest scientific standards”, “to conceptualize the models as ‘knowledge representations'”, “to include the surrounding environmental context”, “to utilize real-time visual simulation”, and “to promote access and utilization by other scholars”. A great part Digital Karnak is the “External Resources” tab for further exploration of the topic of Karnak specifically and Egyptology as a whole.

Week 8: Current Events

This week I explored “The Knotted Line” website. I thought the design was very interesting and the interactivity was key part of my experience. The method of accessing the information relates to the information presented. The timeline presents the history of oppression in America, especially how it relates to education, housing, and prison systems in the more modern era. Sometimes it was a struggle to get the blurb to pop-up which is a strength of the website not a weakness. I think this parallels the difficulties present when trying to access information while oppressed. Sometimes I also had to access a previous pop-up before the next one could show up, which re-enforces the idea that knowing the historical context and previous events is important to understanding our current social climate.

The About section describes how to use the site but if you want to access the background the the project you have to click through one of the blurb pop-ups. Because this is a timeline with imagined future events I would like to know when this project was created. A great strength of this site are the pages you can access once you click through one of the blurbs. Each item on the timeline is listed is chronological order and you can also browse by theme/subject matter. The homepage is a piece of artwork in itself and the “back page” is a great tool to expand on the information presented on the homepage, especially if you missed some items on the timeline.

When I explored the timeline I immediately thought of California’s recently approved Proposition 47. Prop 47 reduces the classification of most “nonserious and nonviolent property and drug crimes” from a felony to a misdemeanor.  Because “The Knotted Line” was created before 2012, the approval of this proposition is not on the timeline but I’m surprised that an equivalent is not present in the “future events” part. The 2016 legalization of marijuana node could be an equivalent, especially because of the focus on the decrease in drug arrests. Prop 47 covers more than drug-related crimes and will lead to a decrease in the number of people that have to go to prison for crimes committed. The “future events” on the timeline focus on prison reformation as a major point and it also focuses on shifting laws that also improve the entire justice system. This is important because it shows that “The Knotted Line” is looking at America’s history of oppression from many different angles and doesn’t consider prison reform as the ultimate solution. Instead it’s a combination of factors that effect the status of our freedoms.

Week 7: Web Maps

I’ll start off by saying: I am not a fan of the common web-maps. I think Google Maps, MapQuest, etc… etc… are horrible and I hate working with them but that’s just my personal opinion. Despite my vendetta, the “Anatomy of a Web Map” was an interesting site. I thought it was interesting how the whole premise in the beginning is “tiles are revolutionary in the web-map field!!!” when they are commonly used in satellite image analysis, which just highlights the evolutionary process and history of mapping. While a map from Google might be divided into “tiles” a satellite map is divided in “scenes” based on the satellite’s orbit around the earth. I had some issues with “Anatomy of a Web Map”, mostly because it took forever to load, but also their emphasis on raster layers, tiles, and base-maps. As someone who has experience making physical paper, digital, and web maps, vectors layers are much easier to work with and base-maps make everything a nightmare. This site really made me aware of my mapping preferences. A map like Google maps, although commonly used, is incredibly weak. Many areas of the world either do not have any information or the information is wrong. A program like Open Street Maps is cool however because you can go in and add missing data and change the map versus accepting the map as truth.

I preferred the “Introduction to Web Mapping” over the “Anatomy of a Web Map” just because it’s more in line with how I like to make and use maps. Both sources can be applied to this great map I found of the most popular professions in American cities according to Linkedin profiles. And there is a map of the most popular professions in European cities on the same page as well. The link to the page is provided below. These maps are great examples of dynamic, interactive maps. The symbology is great, using colors to represent different qualitative data. While the base-map may use tiles to load the symbol layer doesn’t. This is in comparison to Instagram’s method of grouping pictures taken in close proximity more and more depending on the zoom level, which many people have written about. Each point in the symbol layer will be shown at any zoom level. This makes sense because the data is divided by city and isn’t group also by geographic region or state or etc….  This is one reason why I have a beef with tile-based web-maps and base-maps; sometimes the level of their data and dynamics are different to the dynamics of the actual data points.

MAPS!

Week 6: Networks in Literature

A Song of Ice and Fire Network Maps

I’m a huge nerd and I love the Song of Ice and Fire series. The story is great but the connections between characters is very complicated. While capta that come from real-world data is hard enough to map out, the character relationships in the series would be almost impossible. According to Demystifying Networks there are algorithms that help to sort out the “stuff” into nodes and the edge weights between the nodes. Creating a network can work when the “stuff” that gets put into nodes is static and doesn’t change over time. The relationship between nodes would also have to be static. The example provided in the Demystifying Networks article is a good example of static “stuff”. The author of a book isn’t going to change over time (unless it comes out that it was plagiarized from another author) That’s why I think the network of relationships between characters and houses in A Song of Ice and Fire is so problematic. The best part of the narrative is how the relationships between houses and individual characters change so drastically. As the Demystifying Networks article states “data that does not fit neatly into one category or the other” and this is true in the character network map especially when placing characters together according to their houses, which changes due to marriages often.

It would be best to create a network map that also includes a temporal component but then at that point you might as well just read the books. The relationships and network data is already so convoluted and complex that adding a temporal component might over-complicate even more but it would add important information that is currently missing from the network map. While in this case, it might not be important to figure out a way to present this data. But real-life data is just as complex as the data found with books as the main source of data (versus for instance using real-life data as a main source).  Like any way we present information, how in-depth is too in-depth? When does presenting data turn into just regurgitating the data as-is?

Week 5: Capta to Data

Even though Drucker’s article was mostly about how we visually represent knowledge it was definitely about how we represent knowledge as data too. Today I looked at Max Fisher’s map of the world’s most and least racially tolerant countries. A few questions came to mind. The main one being how did he turn the data of racial intolerance into capta? How was he able to take something as hard to define as racism and qualify it into a visual representation? The short answer is, he didn’t even collect his data correctly. Racial intolerance is something that changes from group to group and definitely from country to country.

According to Siddhartha Mitter’s article about Fisher’s map “The Cartography of Bullshit“, Iranians were asked about Zoroastrians; Puerto Ricans about Spiritists; Tanzanians, about witchdoctors, etc… Despite the questions being different the answers were presented as the same. All of the answers to the leading questions in Fisher’s survey I assume got put into an algorithm with the output being a country’s intolerance level. But this data is not something that can put on the nice blue to red color axis we see on his map. It would be much better to use a model similar to Drucker’s bar graph for gender data as both gender and race issues do not fit on a simple binary and it’s misleading when they are presented as such in data visualizations. This type of data cannot be measured on one metric.

The main problem I have with Fisher’s map is that even though it would be incredible difficult to visually represent this data he messed up at the initially step of turning the capta into truthful data. I think that the survey might have been designed a certain way or the data might have been manipulated in order to be able to represent that data in a visual format, specifically the map. Unfortunately with the current state of out understanding of data visualization, a strict textual representation might be the only way we can truthfully represent knowledge that ambiguous classification. But forcing data that can’t be easily defined as capta into a visual format Fisher misled and created a problematic info-graphic that people are going to be ready to believe even though that the data itself on a fundamental level is flawed.

Fisher’s Map

 

 

Data + Design

My Cartography assignment

The Data + Design was super interesting to read and provided a lot of really valuable information. I wish I found it earlier because I could have applied a lot of the information on data presentation to my assignments for a cartography class I’m taking. For the assignment I linked above I did not collect the data myself so I can’t relate it to the “Collecting Data” section of the online book but I did find the “Visualizing Data” and “What not to do” sections very relevant. The style of the online book has that classic mac sleek look to it. In my experience, a lot of websites or info-graphics that do this often over simplify or misrepresent data. However, I feel like this book did a good job in presenting its information. Due to all the different “chapters”I feel that the information got represented fully.
The methods of data collecting were really well explained in both text and visual format. The importance of accurately presenting information through a visual medium is being ever more important. Because visuals are often widely used and understood it’s important to make sure that they portray the correct information and this is mostly controlled by the presenter. In cartography there are four “quadrants” in the cartographic process but these can be applied to any representation of data. The two quadrants determined by the presenter are pattern-recognition and encoding. In general pattern-recognition is deciding what data to use and encoding is deciding how to represent that data. The viewer’s two quadrants are decoding and spatial cognition. Decoding is how the viewer understands the information presented and spatial cognition is how that information effects their view of the world. Although the spatial cognition quadrant is mostly restricted to cartography, the rest of the quadrants can be easily applied to any sort of data.
In the “Perception Deception” chapter the most relevant section for me in terms of my cartography assignment was the small paragraph about the contrast between level differences. Luckily the map-making tool I was using, Google API uses a color gradient assigned to the data values so that level difference isn’t an issue. However, if you want you, can define colors for specific data values if you want in the Google API code. Because this type of map (a choropleth map) is purely based on the color of an area, as the value of the color is the ‘symbol’, it’s super important that the colors are chosen wisely.

Why the Silly Genres?

This week my post focuses on the “How Netflix Reverse Engineered Hollywood” article. Thinking about the n-dimensional classification scheme while reading this article was interesting. The classification Netflix uses isn’t n-dimensional because the genre title can only include so many characters. But it’s not even fourth or fifth or however many identifiers can be put in the genre title. From my understanding, each classifying tag is independent from the next. If you wanted to categorize “Night of the Living Dead” it would be Zombie movie from the 1960’s, but neither “zombie movie” or “from the 1960’s” will be a branch under the other identifier. The Netflix classification scheme is almost an ultra-specific 1-dimensional classification scheme. I looked to the Netflix terms of use to see if their algorithm is mentioned at all. It’s not, but there is a clause that states you may not “engineer or disassemble any software or other products or processes accessible through the Netflix service” which would include their genre system.
However their terms do state that they are constantly updating all facets of their service, which would include their genre list. Based on what people are watching and the growing or dying popularity of a movie or TV show can effect what genres are being included. This can also be applied in the opposite fashion. In the case of House of Cards, Netflix chose to create a Machiavellian political thriller because that was a popular genre. Another clause states that “The availability of movies & TV shows to watch will change from time to time, and from country to country.” The country of origin or setting genre tag is useful to Netflix because they can see if Danish movie are popular in America and therefore should provide American customers with more Danish movies or if they are more popular in Japan and then provide Japanese customers with more Danish movie genres.
It is for this reason that the super specific genre system that Netflix uses not only benefits the viewer but also benefits Netflix itself. If somebody rates romantic comedies 5 stars and coming-of-age movies 4 stars, they will get recommended romantic coming-of-age comedies. From Netflix’s perspective, they gained data that tells them that people who watch romantic comedies also like to watch coming-of-age movies. This data let’s Netflix know the best way to group their movies and TV shows, and what type of movies they should spend money to get licensing for.

“Netflix Terms of Use.” Netflix. N.p., n.d. Web. 20 Oct. 2014. .

Classifications in Practice

I walked into a super market the other day and immediately noticed that the food was divided by type. The vegetables were in one area near the fruit, the dairy had its own refrigerator and the frozen food was kept in two aisles worth of freezers. While most things are hard to classify and even when they are those classifications are arbitrary, it makes sense to classify food because of their differing shelf lives. But even in the simple categorization of food there are some problems and even with “n” amount of categories there are still an infinite amount of categories not included, as mentioned in the McQueen article. A tomato is in the vegetable section but it has seeds like any fruit would, but there was no “vegetables with fruit-like qualities” sign adorning the top of an aisle. Citrus fruit is grouped together but they aren’t arranged by sour-ness. This line of thought is definitely nit-picky and unrealistic but it just shows that there is an infinite number of possible ways to categorize something as simple as food.

The level to which we categorize data and how categorize it effects how our reader interprets said data. When we leave out the infinite amount of categories we are leaving out an infinite amount of interpretations. It’s up to presenters to decide how the data gets organized and what message they want to send. One-dimensional schemes are often the simplest to understand but leave out a lot of information. N-dimensional schemes are the most informative but as n increases the difficulty of categorization increases for the presenter and it is harder for the reader to understand as well. Because categories highlight differences between things the most legible scheme is the simplest one with only two groupings that have clear boundaries, for example: true or false, legal or illegal, 0 or 1. This static grouping is effective when the differences are based in science but become ineffective when analyzing data from the ambiguous humanities realm. Linnaen taxonomy works because it’s based on morphology and DNA but the gender binary is arbitrary because it is based on loose social constructs.

I think this is the central crux of digital humanities. Like the super market, we have to arrange our data so it’s understandable to the reader and can be easily sorted through. But we are presented with a challenge because data from the humanities does not always fit into a nice, neat category. How do we arrange our data and what are we saying with whatever classification scheme we end up using?

Works Cited:  Haliburton, Andrew. Bicycle Needs. 2010. 2010 EVC, Palm Springs. Andrew Haliburton. Web. 13 Oct 2014.