The article “Mortality in the North Dublin Union during the Great Famine” by Timothy Guinnane and Cormac O Grada brings up some interesting points. The article is focused on mortality rates in workhouses in Northern Ireland during the Great Famine. One thing that I noticed in the article is that they only used information form a single workhouse. I don’t believe this is an accurate way to represent mortality rates or an accurate way for people to study this part of history. Using only one workhouse is a very small sample size and there could have been variables at that specific workhouse which weren’t taken into account. The researchers also could have been biased, considering that the topic is so important, and trying to push for change. My classmate JN made a similar point when they said “no data is immune to outside conditions.” The public is not generally involved in the processes of these studies, so how can people blindly trust the information that they provide?

I found this piece on psychohistory by Isaac Asimov to be very intriguing, but at the same time somewhat confusing. Asimov defines psychohistory to be a branch of mathematics which “deals with the reactions of human conglomerates to fixed social and economic stimuli”, but also says that this is accomplished using non-mathematical methods and concepts. This left me confused as to how psychohistory can be a branch of mathematics which doesn’t use mathematical concepts, but Asimov does not explain this much further. I also don’t believe that the story-like writing style of this piece is an effective way to help the readers understand psychohistory. With no background knowledge of what the two people are talking about, it’s hard to follow the conversation completely. But, overall the piece is interesting and opens many interesting questions, such as how much should the public know in regards to the use of psychohistory. The student J-OS said that “the social issues we try to solve are never answered through one set of mathematical data, but rather through the relations and trends between multiple data sets recorded by our society.” I think this is a good way to look at psychohistory and to gain a better understanding of it.

This article by The Future of Research Communications and e-Scholarship discusses challenges and guidelines to creating and using fair data which is accessible and usable by different groups of people. The article lays out a general guideline  using the acronym FAIR: Findable, Accessible, Interoperable and Re-usable. These facets are then explained in deeper detail in the article, allowing people who are not very educated on the use of data to better understand what these guidelines mean and how they are connected. In RC’s response to this article, they said “if all data work/collection follows universal standards it would greatly increase ease of use for every part involved.” The student explained the overall objective of this article and these guidelines very well. Although this quote is much easier said than done, this article and thinking like this are a start.

This reading consists of a thread on Twitter regarding safety and use of data through social medias by John Abowd, the chief scientist of the US Census Bureau. Abowd discusses how data that is made public may not be as safe as it used to be or as many people believe it to be. If too much data is released, there is the possibility of database reconstruction, meaning people or a program could use the data to identify individuals whose names are not meant to be public knowledge. This could be harmful to many individuals who believed that their information was safe and also brings up the question of how much control people have over their data. I like that Abowd is very honest and explains things in a way which are understandable to people with little knowledge on the subject. In a response by JH, the student discusses the positives and negatives of social media data being used by researchers. Overall, I agree with their statements and believe that there is a lot which needs to be changed in regards to social media and these changes should be a larger topic of conversation.  

Topic modeling is a method used by programmers to analyze a large number of documents or text files. The documents are first manipulated to fix possible errors and make it easier for the computer. For example, all words are lowercased, punctuation is removed, and the tense of all words may be changed to a single form. This removes possible human errors in capitalization,  punctuation, and other categories. Meeks and Weingart argue that current topic modeling methods haven’t yet been perfected and for this reason shouldn’t be used in analysis. But, I agree with HC when they say that they “don’t believe we should give up on the whole field just because they haven’t produced a perfect finished product yet,” as there are many things in the computer world which aren’t yet flawless. For example, many companies use artificial intelligence for different aspects of products and business. Although artificial intelligence is nowhere near perfected, it constantly produces useful results for companies and consumers.

I found the article Using Metadata to Find Paul Revere very intriguing and an interesting way to use data. The first thing that the reader might notice is the conversational tone of the writing, as pointed out by HR: “the author wrote it like a class lecture or tutorial.” This keeps the reader engaged and  holds the readers attention throughout the piece. The author looks at different organizations of the time and finds how Revere connects to the organizations as well as how the organizations connect to each other. The author uses visuals, showing connections and strength of connections between organizations based on number of shared members. The visuals are good, but could be better if the user was able to interact with the visuals and possibly even click on individuals or organizations and only see that specific entity’s connections. This use of metadata and networking could be used in present day to learn more about terrorists, their organizations, and how they all relate.

In this article, Lincoln Mullen discusses the practice of lynching as well as the reasoning behind lynching. He also shows and describes multiple visuals that were made to help readers visualize the prevalence of lynchings and to “awaken the public conscience by making the extent of lynching visible.” But, how affective are these visuals? The first visual is a chart made to show the number of lynchings per week by year from 1877 to 1950. But, the chart is unorganized and makes me feel as though I’m looking at a game of Mine Sweeper. The different shades of red are too similar to tell the number of lynchings that are being represented and all of this data is squeezed into too small of a space. The second visual is much more helpful, depicting the number of lynchings per year using a line graph. This graph clearly shows trends, is well organized, and is easy to understand. The third visual is an interactive map of the United States that shows number of lynchings per county. When the reader clicks on a county, they can see the number of lynchings, number of victims, and dates of the first and last lynchings in each specific county. I believe that interactive visuals are a great way to portray information to readers. The visual is another map, but it is less understandable because it is much older and not interactive, so this is understandable and possibly the best map the creators could make at the time. The final two visuals are both graphs of number of executions by race from 1800 to 2002. Although the graphs are similar, I believe that the second graph shows the information much more clearly than the first because it is not filled in and a reader can clearly track the different lines. SJ said “I genuinely believe the visualizations produced did not help with any new or essential trends,” but I disagree. The visuals make these trends accessible and understandable to the public.

In the article “Locating Place Names At Scale: Using Natural Language Processing To Identify Geographical Information In Text”, the authors discuss how similarly named locations are identified using broad context without needing a human to decipher the text. Researchers use Named Entity Recognition (NER), which is a method that computers can use to process language, to locate names of places within text. They then start to find references to cities, states, or countries using Google API to narrow down the possibilities until there is only one left. If there is more than one possibility, whichever place is closest to the reference points is deemed correct. As DA put it, this article “explore[s] the fault in some programming” and data which is  a result of these faults. The article also demonstrates a way that people were able to circumvent these faults and create a solution. 

Welcome to WordPress. This is your first post. Edit or delete it, then start blogging!