ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Dataset NameShort description of datasetLink to data source (where you downloaded it)Link(s) to documentation of data and/or projectNumber of recordsWho collected the data or created the dataset?Who owns or holds the data, collection, and/or original materials the dataset is based on?How was the data collected?Who/what is included and/or excluded from dataset?When was the data collected or dataset published/released?Why was the data collected?Who is/are the audience(s) for this data?Is there anything you want to know about this dataset or how it was created that you can't find out (or it's not clear if you can find out) from the dataset itself or its documentation?Notes
2
visual-style-shot-dataThe dataset is a collection of shot data and information for use in Arnold and Tilton's larger project on Distant Viewing. The project is centered around facial recognition software. Arnold and Tilton have developed a computational methodology, distant viewing, as a form of machine learning that is capable of extracting "shot semantic" information (such as editing techniques, visual composition, framing) from the number of faces that appear within a given shot and the proximity of those faces to the camera. The goal is to extract imagery and facial placement to learn more about visual construction.https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/S84TSXhttps://culturalanalytics.org/article/11045-visual-style-in-two-network-era-sitcoms102333Taylor Arnold, Lauren Tilton, and Annie BerkTaylor Arnold (University of Richmond)Arnold and Tilton compiled a number of episodes from two series, Bewitched and I Dream of Genie. Each episode was then individually divided into a sequence of shots and frames. Frames are the individual images which, when played in sequence, produce the illusion of seamless movement. The concept of 24 frames a second, or 60 frames a second, reveals the number of frames played in succession per second. More frames is equivalent to a smoother image. Most films and television shows are shot at 24 frames per second. These frames are then feed into a computer, along with facial reference information, in order to determine position of faces, composition, and framing. The dataset is compiled of shot compositions from two series, Bewitched and I Dream of Genie. Both series began airing in the mid-sixties (1964 for Bewitched, 1965 for I Dream of Genie) and involve a predominatly white heto-normative cast. This is not, in itself, intolerable but it does call into question the utility of facial recognition software. There is ample evidence of computer facial recognition software (such as the facial tracking on certain web cameras) being unable to track and monitor black and brown peoples faces. It is, then, interesting that Arnold and Tilton have opted for two series with white cast members, and make no reference to this exisiting limitation in facial recognition software.The data appears to have been compiled over the course of 2019. It seems that much of the data was put together over the course of April and published in August concurrently with "A Visual Style in Two Nework Sitcoms".In answering this question, I would like to turn to Arnold and Tilton's article, "A Visual Style in Two Network Sitcoms". In this paper, Arnold and Tilton conceptualize the distant viewing project, and the specific functionality of facial recognition software in narrative and culutral analysis, as "how face detection and recognition algorithms, applied to frames extracted from a corpus of moving images, can capture formal elements present in media beyond shot length and average color measurements. Locating and identifying faces makes it possible to algorithmically extract time-coded labels that directly correspond to concepts and taxonomies established within film theory. For example, knowing the size of detected faces, for example, provides a direct link to the concept of shot framing. The blocking of a scene can similarly be deduced by knowing the relative positions of identified characters within a specific cut. Once produced on a large scale, these extracted formal elements can be aggregated to explore visual style across a collection of materials. It is then possible to understand how visual style is used within the internal construction of narrative and as a way to engage broadly with external cultural forces." (2). In other words, the data was collected to provide an empirical, testable, and computational methodology that can support new and existing "concepts and taxonomies" in film theory. Based on my analysis of the dataset, the distant viewing model appears geared towards film theorists. Specifically, it is an attempt to empirically ground film theory in camera positions and shot compositions in relation to facial recognition software sensitive to the size of faces and the number of faces that appear on screen. To reiterate a point I made earlier, I believe this facial recognition software must be used on non-white actors in order to test the viability of this software. Further, on a formal level, it would be interesting to see the applicability of the distant viewing methodology on the so-called "prime television" of the 2000s. Network television, especially of the two camera set-up of most situational comedies, carries a distinct visual style. How does distant viewing provide usable metadata for shows as stylistic distinct as, to name a few examples, FX's Legion, HBO's The Leftovers, and AMC's Mad Men?In the fields or variables sheet I have provided the data for a small sample of the existing observations. It would be simply unnecessary to include the entire dataset, as it consists of over 100,000 observations. The sample consists of fifteen observations from both Bewitched and I Dream of Genie.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100