<!--[if gte mso 9]><![endif]--><!--[if gte mso 9]><![endif]-->
<!--[if gte mso 9]><![endif]--><!--[if gte mso 9]><![endif]--><!--[if gte mso 9]><![endif]--><!--[if gte mso 9]><![endif]--><!--[if gte mso 9]><![endif]-->
PRIMARY
Gustavo
Dejean, Universidad Nacional del Oeste – Universidad de Buenos
Aires, dejean2010@gmail.com
Student Team: YES
Did you use data from both mini-challenges? NO
Tableau; IBM SPSS Statistics v22; JMP 11 (SW), PostgreSQL
Approximately how many hours were spent working on this submission in total? 120 hours
May we post your submission in the Visual Analytics Benchmark Repository after VAST Challenge 2015 is complete? YES
Video Download
Video:
Example:
https://www.youtube.com/user/gustavoDejean/videos
PART 1
Questions
MC1.1 – Characterize the attendance at DinoFun World on this weekend. Describe up to twelve different types of groups at the park on this weekend. Limit your response to no more than 12 images and 1000 words.
Type Group 0 (without check-in)
a. How big is this type of group?
From 1 to 8 persons, totaling around thirty-nine (39) people. totaling around 490 persons.
b. Where does this type of group like to go in the park?
This group records only the Entry check-in. Overall, the group moved through the entire park. The group attended the park for the three (3) days or only for one (1) day (No Two-Day attendance).
c. How common is this type of group?
There is no other group with such features. These are the Point Outliers of the Linear Model shown in Figure 1.
d. What are your other observations about this type of group?
Linear Correlation and Outliers
A linear correlation was identified between the movement count and the check-in count. More accurately, 84.8% of check-ins may be inferred from the movement count. This is a very good correlation that may be used for identifying any anomalous case that does not meet such correlation. It may also be used for finding types of groups with common features. Figure 1 shows the Linear Model, where anomalous cases without any check-in (39 cases) may be seen on the X-coordinate. Such cases are divided into three sub-groups, according to hours of permanence and movement count.
e. What can you infer about this type of group?
Very likely, they are employees from the Park Security, or Sanitation or Surveillance departments.
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?
No Improvements
Figure
1.1. shows the linear correlation existing between the move count and
the check-in count.
Color intensity shows hours of permanence
inside the park as measured in hours. IDs without Check-Ins are on
the X-Coordinate. They were divided into 3 subgroups, according to
hours of permanence and movement count. Please
note that although each point represents an ID, the IDs of a group,
in general, are very close from one another and, in most cases, are
overlapped. Overall, if both scales are enlarged, IDs may be better
viewed; and, conversely, if both scales are reduced, it is the groups
that are better viewed.
Type Group 3 (extreme gamers)
a. How big is this type of group?
This type of group is mostly made of 1 to 6 persons, but there are twenty groups having between 7 and 31 persons. (at least one of them is a sub-group deriving from another large group).
b. Where does this type of group like to go in the park?
This type of group prefers games like Thill Rides (See Figure 1.6) while they hate Kiddie Rides (See Figure 1.7). These are the Outliers of three out of the four box graphs shown in Figure 1.4.
c. How common is this type of group?
There are approximately 174 groups of this type. By adding up all their individuals, they stand for 5.35% of the overall visitors. (606 people).
d. What are your other observations about this type of group?
This type of group is the one having the highest rate of move counts and check-in counts. They attended the Park for either two or three days. Visually, they are located in the furthest top right of Figure 1.3 and they form part of those that are outside the Normal Bi-Variant ellipse 0.9. In Figure 1.4, they are the Blue Outliers in the three box graphs. These groups belong to Cluster 3, in Figure 1.2. In the Figure 1.5 are in colour green (cluster 3).
e. What can you infer about this type of group?
There are no children in this type of group. This group members enjoy a lot games such as Thrill Rides. They move much more than average visitors. This may be seen in Figure 1.3; they are well above the straight line in the Linear Model. They must be groups of youngsters.
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?
These groups are ideal for receiving a second invitation to the park as well as advertisements with special offers. Their loyalty should be assured.
Figure 1.2. shows the best result obtained with K-mens. The 39 missing pieces of data are those that were segregated in Group 0.
Figure 1.3
Figure 1.4. There is a relationship of rejection between the Kiddie Rides (KR) and the Thrill Rides (TR): please note in Blue that the Outliers of the TRs are those that have a 0 in KRs (dark green).
Figure 1,5
Figura 1.6
Figure 1.7. Large differences are observed between the types of groups belonging to different clusters
Type Group 5 (colour orange Figure 1.5)
a. How big is this type of group?
From 1 to 4 persons, with few exceptions going from 5 to 62 persons.
b. Where does this type of group like to go in the park?
This type of group prefers KR games than types of groups 1 and 3 (See Figure 1.8) and may be compared with type of group 4 in their liking of TR games (See Figure 1.9).
c. How common is this type of group?
There are approximately 909 groups in this category.
d. What are your other observations about this type of group?
They have a low rate of move counts and check-in counts. This is the most regular type of group and the most opposite to type 3. Generally, they attend the park only for one day. Compared with groups 1 and 3, they like Kiddies Rides more (See Figure 1.8); (as they attend the only for one day, the average per day is compared here). In the Figure 1.5 are in colour orange (cluster 5).
e. What can you infer about this type of group?
In general, this type of group is made up of children. As they know they will attend for only one day, they try to manage their time correctly and allot consistent time to all the different entertainments.
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?
No Improvements
Figure 1.8
Figure 1.9
Type Group 6 (like shows + KR)
a. How big is this type of group?
From 1 to 7 persons, with few exceptions going from 5 to 27 persons.
b. Where does this type of group like to go in the park?
In average, they show a high rate of participation in games KR (See Figures 1.7 and 1.8) and they are the most frequent participants in Shows. (See Figure 1.10)
c. How common is this type of group?
There are approximately 176 groups of this type, totaling around 490 persons.
d. What are your other observations about this type of group?
They generally attend for three days. Recorded many check-in entry (ver f1.11). In the Figure 1.5 are in colour turquoise (cluster 6).
e. What can you infer about this type of group?
This group has children.
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?
No Improvements
Figura 1.10
Figura 1.11
Type Group 1
a. How big is this type of group?
From 1 to 13 persons, with only 9 exceptions going from 18 a 49 persons.
b. Where does this type of group like to go in the park?
This type of group prefers games like Everyone and Thill Rides (See Figure 1.6) while they hate Kiddie Rides ( ver F1.8 y 1.9).
c. How common is this type of group?
There are approximately 371 groups, totaling around 1407 persons.
d. What are your other observations about this type of group?
They generally attend for one or two days. (Figure1,12)
e. What can you infer about this type of group?
Moving more quickly then normal. There are no children in this type of group. In the Figure 1.5 are in colour blue (cluster 1).
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?
No Improvements
Type Group 2
a. How big is this type of group?
From 1 to 8 people the mayority , but with more than 50 exceptions that are 10 to 58 people .
b. Where does this type of group like to go in the park?
It has preferences for the games Kidder Rides.
c. How common is this type of group?
There are approximately 347 groups, reaching 2488 people.
d. What are your other observations about this type of group?
Generally, they attend the park only for one day ( see figure 1.12). It is a type of group made up of bigger groups. (cluster 3 colour red, figure 1.5)
e. What can you infer about this type of group?
They have children.
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?O
Offer transport service in the park.
Type Group 4 (violet)
a. How big is this type of group?
From 1 to 9 people the mayority , but with 2 exceptions; one with 20 and other with 24 persons.
b. Where does this type of group like to go in the park?
Shows and Kidder Rides
c. How common is this type of group?
There are approximately 906 groups, reaching 2305 people.
d. What are your other observations about this type of group?
They generally attend two or three days. (see figure 1,12). In the Figure 1.5 are in colour violet (cluster 4), in the center graph.
e. What can you infer about this type of group?
Moving more slowly then normal. They have children.
f. If you were to make one improvement to the park to better meet this group’s needs, what would it be?No Improvements
Figure 1,12
Part 2
MC1.2 – Are there notable differences in the patterns of activity on in the park across the three days? Please describe the notable difference you see.
Limit your response to no more than 3 images and 300 words.
Answer
MC1.2.1 Closing of Shows & Entertainment 32 and 63 on Sunday
Figure 2.1. shows items 32 and 63, which were closed to the public from 12:00 on Sunday, as no further check-ins were recorded from that time onwards. This evidence is taken into account for inferring that the robbery was discovered on that day and time, and during Item #32. (day 8 - 11:59:00 h. very approximately in the Creighton Pavillion). Besides, the great difference between Fridays and Saturdays vis a vis Sunday may be observed.
Figure 2.1 shows check-in distribution for Shows 32 (Blue) and 63 (Orange), in each one of the three days (6, 7, 8). On Sunday, from 12:00 onwards, no further Check-Ins were recorded in any of these two Shows.
MC1.2.2 A decrease in the audience only in three Sunday entertainments.There are only three items having decreases in audience in absolute terms on Sunday. Item 32 (as described above); item 63 (Shows), and item 4 (TerrorSaur, in the Thrill Rides) (See Figure 2.2).
Figure 2.2 shows the number of check-ins for each one of the entertainments throughout the three days. There are only three negative inflections: two in red for Shows (Items 32 & 64), and one for Terror Saur (Item 4). Each color stands for an Item.
MC1.2.3. Higher demand at Information Desks on Fridays
Check-ins of at Information & Assistance (Item 62) have a dramatic increase on Friday, from 13:00 to 15:00. (See Figure 2.3).
Figure 2.3. shows the number of check-ins in 62 for the three days, divided per hour.
Part 3
MC1.3 – What anomalies or unusual patterns do you see? Describe no more than 10 anomalies, and prioritize those unusual patterns that you think are most likely to be relevant to the crime.
Limit your response to no more than 10 images and 500 words.
MC1.3.1 ID 1620233 (Main Suspect)
Recorded three check-ins on the day and location of the robbery: at 11:39, at 11:51, and at 11:59. Its last movement was recorded at 13:20:22. Unusual aspects and coincidences for this ID are as follows: a) Being at the same place and time of the robbery; b) An unusually early exit; c) Exiting after robbery time; d) Three check-ins at the location of the robbery (the only case). It acts solo.
MC1.3.2 ID 1983765 features an overlapping: two different and simultaneous itineraries are recorded
On Day 7, at 20:18, ID 1983765 records simultaneous movements in two different far apart points starting in: (76, 36) and (31, 34), until such movements reunited at the Entry (0, 67) at 20:34, and then, exited. This to demostrate that you have the possibility to simulate their movements. (See Figure 3.1 )
Figure 3.1 Intineraty of the id 1983765, can evade the control
MC1.3.3 Two groups that exit unusually earlier, following the Main Suspect exit
Two groups that exit unusually earlier on the day of the robbery, only some minutes after its perpetration.
MC1.3.4 ID 657863 stayed overnight.
The group made up of IDs {657863 , 103006, 313073,1412235, and 1937834} had an anomalous behavior: all of these IDs entered at the same time; they shared the Day 6 Itinerary including a stop at Food Store 39. The last four exited the park on the same day, but the first one stayed on in Entertainment 20. On the next day, it left at 8:00 am and exited the park at 08:10 am. (See Figure 3.2).
Figure 3.2 shows the recordings of movements of ID 657863 on Friday 6 (Blue), and on Saturday 7 (Orange). On Friday, this ID checked-in at 04:11:36 pm, and its last movement was recorded in Item 20, at 09:53:11 pm. On Saturday, it left Item 20 at 08:00:27 am, headed to the Entry Corridor and arrived there at 08:10:03 am.
MC1.3.5 Enter late and exit quickly
This grup is made up of the following IDs: 195087, 1614660, and 2096515. What strikes us about this group is that it entered at a late time and it exited two hours later, without any check-in activity being recorded inside the Park. Figure 3.3 shows the itinerary.
MC1.3.6 ID 1600469 recorded a quick relocation.
On day 8, at 9:00, ID 1600469 recorded movements at a constant speed from the Entry Point (57, 66) to (23,20) at 19:18. The speed rate was 10 m/second, an indicator that such relocation was done in a vehicle. This ID is one of the Outliers of the Linear Model and has already been included in the Security personnel group.
MC1.3.7
Registran 12 entradas ( cuatro entradas por dia). Tienen cantidades de juego parecidos y comparten horas de llegada y de partida.