Difference between revisions of "Wyz-Visualization & Insights"
Line 117: | Line 117: | ||
<td>[[File:Wyz-flu-related microblog messages symptom.png|800px|center]]</td> | <td>[[File:Wyz-flu-related microblog messages symptom.png|800px|center]]</td> | ||
<td><b> Changes in human traffic during the day and night</b> | <td><b> Changes in human traffic during the day and night</b> | ||
− | <br> | + | <br>What is also noticeable is that the areas affected spread to the downstream of Vast river on May 19th. What’s more, we discovered that the posts related to gastrointestinal problems increased gradually. The bar chart gives a breakdown of the different symptoms reported in microblog from May 17th to May 20th. It is apparently manifest from the graph that the posts related to gastrointestinal issues experienced a dramatic increase over the period from May 19th to May 20th. From the map, the reports are dense in the lower reaches of the Vast river. |
+ | </td> | ||
+ | </tr> | ||
+ | <tr> | ||
+ | <td>[[File:Wyz-wordcloud.png|800px|center]]</td> | ||
+ | <td><b> Wordcloud by term frequency </b> | ||
+ | <br> The word cloud shows the flu-related words with high frequency in microblogs of 19th and 20th. Highlighted words "stomach" and "diarrhea" which describe abdominal problems hadn't appeared until 19th. | ||
+ | </td> | ||
+ | </tr> | ||
+ | <tr> | ||
+ | <td>[[File:Wyz-diarrheatext.png |800px|center]]</td> | ||
+ | <td><b> Wordcloud by term frequency </b> | ||
+ | <br> To go further about this issue, the messages containing abdominal problems related words are extracted, which is shown in the screenshot to see the details. | ||
</td> | </td> | ||
</tr> | </tr> | ||
</table> | </table> |
Revision as of 16:37, 14 October 2017
|
|
|
|
Visualization & Insights
Contents
Origin and Epidemic Spread
Identify approximately where the outbreak started on the map (ground zero location). Outline the affected area. Explain how you arrived at your conclusion.
After a quick browse around the dataset, we can find that the content of over 1M microblog messages contains all aspects of life. So first, we start by extracting flu-related microblog messages. A list of keywords was selected to filter raw data. The list consists of keywords including flu, chill, fever, sweat, ache, pain, fatigue, cough, breath, nausea, vomit, diarrhea, and lymph, which are from observed symptoms and human judgement. After filtering, there are 71939 flu-related microblog messages left.
Visualization | Insights |
---|---|
Infer the disease outbreak date
|
|
Map of flu-related microblog messages on May 17
|
|
Map of flu-related microblog messages on May 18
|
|
Map of flu-related microblog messages on May 17
|
|
Magnification of ground zero location, May 18
|
Epidemic Spread
Mode of transmission
Present a hypothesis on how the infection is being transmitted. For example, is the method of transmission person-to-person, airborne, waterborne, or something else? Identify the trends that support your hypothesis.
Person-to-person
Considering that most of the flu is spread by person-to-person, we will focus on this route of transmission first.
Visualization | Insights |
---|---|
Number of flu-related microblog messages breakdown by hour, May 18 2011
|
|
Map of flu-related microblog messages by work status
|
Airborne
Visualization | Insights |
---|---|
Changes in human traffic during the day and night
|
Waterborne
Visualization | Insights |
---|---|
Changes in human traffic during the day and night
|
|
Wordcloud by term frequency
|
|
Wordcloud by term frequency
|