Difference between revisions of "ISSS608 Assign3 HoLiChin Task2"

From Visual Analytics and Applications
Jump to navigation Jump to search
Line 79: Line 79:
  
 
===Pattern #4 – Communication Patterns at Creighton Pavilion===
 
===Pattern #4 – Communication Patterns at Creighton Pavilion===
 +
As some abnormal things happened at Pavilion, next more analysis will zoom in on the communication patterns at Pavilion, to detect the most likely suspect.
 +
 +
From Figure below, it’s observed the communication volume increased starting from about 11:30AM, then 11:32AM, finally peak at 11:44AM.
 +
 +
Now, we first zoom in to examine the patterns between 11:30am to 11.50 am (first peak).
 +
 +
[[File:Q2d.PNG|480px|center|Fig4]] <br/>
 +
 +
A series of things happened at Pavilion on Sunday between 11.30AM to 11.52AM:
 +
* A peak of communication occurred around 11:44AM.
 +
* The top senders sent at Pavilion were highlighted in the rectangle
 +
* Most messages were sent to External during this period.
 +
 +
Another peak of communication occurred around 12:00PM (noon). And now most of messages were sent to 839736. (See figure below)
 +
 +
[[File:Q2e.PNG|480px|center|Fig5]] <br/>
 +
 +
 +
In conclusion, at Pavilion, the peak of communication related to external appeared 10 minutes earlier than the peak of communication related to 839736 at noon on Sunday. There was an increase of messages to external probably due to vandalism that took place, and when police arrived at Pavilion during that time, and that caused visitors to communicate to external for “breaking” some news to outside the park.
 +
 +
There is an increase in messages to the 839736 could probably due to visitors started to contact the Info Center / Park Help desk to enquire on what had happened.
  
 
===Pattern #5 – Detecting the Most Likely Crime Suspect===
 
===Pattern #5 – Detecting the Most Likely Crime Suspect===
  
 
===Pattern #6 - Communication data at Grinosaurus Stage===
 
===Pattern #6 - Communication data at Grinosaurus Stage===

Revision as of 01:54, 28 October 2016

Introduction

Data Preparation

Task 1

Task 2

Task 3

Visualisation Links & Conclusion

 



Task Requirement

Describe up to 10 communications patterns in the data.
Characterize who is communicating, with whom, when and where.
If you have more than 10 patterns to report, please prioritize those patterns that are most likely to relate to the crime. Note: Please limit your response to no more than 10 images and 1000 words

Communication Patterns

Pattern #1 – Big Groups

Fig1


Analysis and Findings

From the above Qlik Sense visualisation, besides ID 1278894 and 839736, there were some IDs top in sending large volume of messages (more than 3000 texts, but less than 5000) over the 3 days. The top 5 IDs (in descending order) are 1045021, 1116329, 1749109, 918738 and 1388162. In the bar chart in lower quadrant, it shows that those high volume IDs had sent messages to a big group of unique receiver IDs.

For example, by drilling down the first ID 1045021, it’s observed it had sent 3.81K messages to about 2.77K unique Receiver ID. This ID was in the park for all 3 days, and most messages were sent on Friday & Sat, at Wet Land. From the above, we might deduce that this group of IDs could be the group leaders sending the bulk messages to the group members, or they could be the Park staff that sent information to visitors.

Pattern #2 - Identifying the Dynamics of Park Activities

Fig2


Analysis and Findings

Some of the observations from the above visualization: -

  • The openings hours of the park were from 8:00am to 23:00pm.
  • Saturday and Sunday had higher communication volume than Friday, that means more visitors over the weekend.
  • Comparing movement and check-in, larger communication happened when visitors checked in at attractions in the park during Saturday and Sunday.
  • Again, similar to earlier analysis, there was a spike of communication data observed for both check-in and movement on Sunday @ about 12pm.
  • From the charts of Sender and Receiver activities during the 3 days, the volume of communication of Thrill rides was much higher than other types on both Saturday and Sunday.
  • One thing to observe is the comm pattern at Entrance followed a periodical pattern of high volume at regular intervals, then the comm ceased at the alternate intervals. This could due to some fixed broadcast information or interactive games that sent by the park staff with their visitors near entrance area. Again an obvious high communication spike observed on Sunday about 12 noon, at Entrance.

Pattern #3 - Abnormal Communication Volume from ID:839736 on Sunday

From Task 1, we have found out ID 839736 stand out with large communication volume on Sunday. Over here, we will further investigate what could had caused the large communication volume on that day.

Fig3



Analysis and Findings

The obvious abnormality in communication volume was observed on Sunday, with two obvious peaks. One very obvious peak was at 12 noon, and another was at about 2pm.

From the Park Map, it showed that most communication sent out by ID 839736 were received by Receiver IDs at Creighton Pavilion, during the peak at 12noon. From the Receiver IDs table, the top 5 IDs who received most messages from 839736 were 1092525, 1601276, 38945, 2013094, 95112.

As given in the Park website, there was a weekend tribute to Scott Jones (renowned football star). In addition, a show of memorabilia of his awards, trophies, and the Olympic Gold medal would be displayed in Creighton Pavilion. However, the event at Pavilion did not go as planned. The display of Scott Jones’s soccer memorabilia in the Creighton Pavilion was vandalized. This could explain why the peak communication volume happened at Creighton Pavilion, as this is where the crime had taken place.

Detailed analysis will be done in Task 3, to investigate the abnormal communication volume happened at Pavilion, and to find out when the vandalism.

Pattern #4 – Communication Patterns at Creighton Pavilion

As some abnormal things happened at Pavilion, next more analysis will zoom in on the communication patterns at Pavilion, to detect the most likely suspect.

From Figure below, it’s observed the communication volume increased starting from about 11:30AM, then 11:32AM, finally peak at 11:44AM.

Now, we first zoom in to examine the patterns between 11:30am to 11.50 am (first peak).

Fig4


A series of things happened at Pavilion on Sunday between 11.30AM to 11.52AM:

  • A peak of communication occurred around 11:44AM.
  • The top senders sent at Pavilion were highlighted in the rectangle
  • Most messages were sent to External during this period.

Another peak of communication occurred around 12:00PM (noon). And now most of messages were sent to 839736. (See figure below)

Fig5



In conclusion, at Pavilion, the peak of communication related to external appeared 10 minutes earlier than the peak of communication related to 839736 at noon on Sunday. There was an increase of messages to external probably due to vandalism that took place, and when police arrived at Pavilion during that time, and that caused visitors to communicate to external for “breaking” some news to outside the park.

There is an increase in messages to the 839736 could probably due to visitors started to contact the Info Center / Park Help desk to enquire on what had happened.

Pattern #5 – Detecting the Most Likely Crime Suspect

Pattern #6 - Communication data at Grinosaurus Stage