Entry Name:
"HKUST-Siwei-MC2"
VAST Challenge
2015
|
|||||||||||||||||||
|
Fig. 1.1 Relationship matrix showing
interactions between visitors |
Fig. 1.2 Network showing call ins and outs |
|
Fig 1.3 Huge connections of in calls and outcalls towards one
person demonstrating the person’s uniqueness |
|
|
To discover the
communication pattern, we developed our system ParkVis.
It contains the Group View (Fig1.1), where each stacked bar represents a
visitor of a group, and the MDS View (Fig. 1.2), where all visitors are
mapped as circles. In our system, if we filter out the population by large volume of
calls, there is one dot at the upper right corner left in the MDS View and
there are 2 IDs 1278894 and 839736 in it. Specifically, they had called
3830254 times and 121630 times in three days respectively. These two people
are not found in MC1 and stayed in Entry Corridor in all three days.
Interestingly they had in-message and out-message with the majority of
people in the park without communicating with each other (Fig 1.3). We
infer that they are park officers and probably responsible for Customer
Services. As a result, the whole picture provides a clear identity of the
people. They have sent and received messages from most people in the park
in three days. And we can hypothesize that they are working for a
particular position, which is about processing in-coming messages. |
|
MC2.2 – 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.
Limit your response
to no more than 10 images and 1000 words.
|
Firstly, trends in communication volume can discover timing of
some critical events. For example, from 10:45 - 11:00 and 15:45 - 16:00 on
Friday and Saturday, there is a peak in terms of volume of communication. We
suspect that during that time the football star was on stage and there were
interactive activities. On Sunday however, there is no such observation.
From the news we know that the event was cancelled after the crime
happened. Therefore we can infer that the crime happened on Sunday.
Fig. 2.1 Peak traffic of communication
helps locating the timing of crime |
|
Secondly, volume of communication are
somewhat periodic, and it means visitors tend to pick up their phones and
give calls at some time. Thus, this pattern demonstrates the dynamics of
the park, and can be attributed to the activities organized by the park.
Fig. 2.2 Periodicity of communication
helps locating the dynamics of park activities |
|
Thirdly, there are a group of 13 people who only communicate with
the two park officers with ID 1278894 and 839736 (see in the MDS view).
They do not communicate with each other. They have no check-ins at any
rides and walk around in the park on Sunday. Sometimes they stayed at one
ride for a long time without check-in. Given that they have the same number
of incoming communication records and outgoing communication records, we
can be more convinced that this group of people were very likely to be
working for the park. |
|
Fig.
2.4 Connection of communication of a group of 13 people only communicated
with the staff in Fig. 1.1 |
|
Forthly,
we have discovered a group of 32 people whose major communication is
inner-group communication. For outer-group communication, they only talked
with two officer IDs 1278894 and
839736, and external.
Fig. 2.4 A group of 32 people who “only” had inner-group
communicated. |
|
|
|
This
is a group of 5 people. In terms of outer-group connections, only one of
them have connections the park officers. This group has no inner-group connections.
|
MC2.3 – From this data, can you hypothesize when the
crime was discovered? Describe your
rationale.
Limit your response
to no more than 3 images and 300 words.
|
|
Fig. 3.1 Abnormal traffic
in Pavilion on Sunday identifies the place of the crime |
Fig. 3.2 Path of ‘body guards’ who only
patrolling in the park revels incident happened in the morning
Fig. 3.3 Suspect ID 921888 (green dot with
line) who contains no riding records and only external communication in the
morning |
|
|
In
pattern observed in the previous section (Fig. 2.1), we highly suspect that
the vandalism was discovered on Sunday some time between 11am and 16pm. The
reason is that on both Friday and Saturday, there were surge of
communication from 10:45 to 11:00, from 11:45 to 12:00 and from 15:45 to
16:00. We hypothesize that the football star came to stage in the morning
and he came back again in the afternoon. We suspect that during the above
time slots, there were interactive activities when the football star is on
stage through the DinoFun App. On
Sunday, however, although the pattern remains unchanged in the morning,
there was no peak of comparable magnitude in the afternoon. We hypothesize
that the vandalism was discovered between 11am and 16pm on Sunday. After
the discovery of the vandalism, the scheduled event was affected. To
further narrow down the possible timespan of the discovery of the
vandalism. We can refer to the following Figure 3.1 ,
which shows that after 12:30 on Sunday there were few people going to
Pavilion. Our
hypothesis can further be consolidated by the path of the body guards of the football star. As shown below in
Figure 3.2, this group of people went to the stage twice on both Friday and
Saturday, once in the morning and once in the afternoon. But on Sunday,
they went to the stage only in the morning. For
suspects, we filter out the people who did use the facilities in the park,
and find out that there were a few people who have extremely low usage of
communication but when they did, those calls were directed to external.
Therefore, combining the time, place and suspicious activities. We can
identify all, if not most, the attributes of the crime. |
|
