Wallack and Srinivasan argue in their paper that the divide between states’ and communities representations of ontologies can lead to information loss that silences the voices of individuals within the decisions made by states. An interesting comparison of two datasets from the L.A. Controller’s Office, shown as the top and the bottom result from the list of highest rated datasets, demonstrates how the differences between the most and the least valued datasets convey the presence of mismatched ontologies between the state and individuals.
The most highly rated dataset, “Balance of All City Funds”, provides 42 data types including the balance of the funds, the source of the funds etc. The record in this dataset is a specific city fund. The dataset on the opposite end of the spectrum, “Neighborhood Council Expenditures for Fiscal Year 2014”, provides only 8 data types like the name of the neighborhood council and the amount of the expenditure. The record in this dataset is a specific expenditure related to a neighborhood council.
Although both datasets can be defined as meta ontologies from Wallack and Srinivasan’s argument, as both are produced by the state of Los Angeles, the first dataset leans towards an inclusive meta ontology while the second dataset embodies a state meta ontology that “sheds much of the local context in order to ensure tractable management” (Wallack, Srinivasan, 2). “Balance of All City Funds” provides detailed information on the sources and purposes of the funds and how they are used eligibly. For example, row 17 and 18 record two funds with the purposes of “Zoo improvement projects” and “Animal shelter facilities” respectively. One can easily combine the two specific categories into one such as “Animal protection”, which is exactly what have been done in the second dataset. The task of each expenditure in “Neighborhood Council Expenditures for Fiscal Year 2014” is labeled with general attributes such as “Office” or “Event”.
The difference in intricacy of the two datasets may account for the difference in their ratings by individuals. Ones who interact with the second dataset can feel isolated from the actions of the state without knowing what exactly are happening behind the curtain, while those who browse through the first dataset can draw a detailed picture of the state’s decision-making. Although the first dataset does have data types like “Grant Receivable Asset”, everyone with or without a speciality in accounting or finance can grasp at least a piece of information from the dataset to understand the budgeting of the city, albeit to varying degrees. The second dataset, however, excludes anyone without the knowledge regarding the specific usage of each expenditure since it does not expose any detail.
I cannot think of any improvements over the ontology of the first dataset, but I see the ontology of the first dataset as a framework to which that of the second dataset should assimilate. From the perspective of an individual who would like know the details regarding an expenditure, if a contact information of the person in charge of the expenditure or the documents related to the expenditure are provided as in the first dataset, he or she can can gain access to the details even if not many numbers or descriptions are record directly in the dataset.
Wow! I never thought to compare my dataset to another one, but this is great. It’s interesting how they are both meta ontologies, but one is easier to read than the other. I also looked at a meta ontology where there was so much detail I’m almost sure no citizen takes the time to read all of this, and if they did, probably wouldn’t understand it fully. I think the that the state could benefit from making these datasets clearer, because, like you said, if one doesn’t have the knowledge regarding the expenditures or terminology, they will get lost.
I really love this blog post! Your comparison and resulting discussion of the two datasets to demonstrate the discrepancy between the way in which meta ontologies are able to be received by the community is perfect in showing the reality of the “digital divide” Wallack and Srinivasan wrote about. I think this showed a great understanding of the essay on your part as well as helped me understand more as well! Thank you!
What a read! Your comparison makes the different ontologies of the 2 datasets demonstrate the dichotomy between the perspectives of the state and communities. Contrary to “City Funds”, “Neighborhood Expenditures” offers few data types and thus limited details about the spending. I agree with your critique on the second dataset, and believe that more description, categorization and explanation are to be added.