Python代写 | Event Detection and Health Analytics Course work

本次英国代写主要为Python网络科学相关的coursework

Course work
• Design an algorithm to detect events from such groups.
• What role entities play? How would you reduce the cost of detecting entities? [4 marks]
• If you were to use tf-idf concepts, how would you capture them? [3 marks]
• How do you remove noisy or spam groups? [3 marks]
• How would you identify categories of events? [3 marks]
•
• Assuming that you have developed a full-fledged event detection system, how
would you show the effectiveness of your detection system? Unfortunately, you
have not test collections to use. Design a crowdsourcing-based design to capture
ground truth and show the effectiveness of the system. [Hint: discuss collecting
data; generating ground truths; what level of annotation process will be used?
What measures will be used for comparison and why?]

• [12]
• Design an algorithm to detect events from such groups.
• What role entities play? How would you reduce the cost of detecting entities?
[4 marks]
• If you were to use tf-idf concepts, how would you capture them? [3 marks]
• How do you remove noisy or spam groups? [3 marks]
• How would you identify categories of events? [3 marks]
• What is an entity
• Examples of named entities include Barack Obama, New York
City, Volkswagen Golf, or anything else that can be named. Named entities
can simply be viewed as entity instances (e.g., New York City is an instance of
a city).

How do we know an entity is present in a text
• SPARQL query to DBPedia
• SELECT ?label ?thumbnail WHERE { <http://dbpedia.org/resource/Harry_Potter>
dbo:abstract ?label . FILTER (lang(?label) = ‘en’)
<http://dbpedia.org/resource/Harry_Potter> dbo:thumbnail ?thumbnail . }
• Limited number of queries is allowed
• Alternative is use an NER model
• Stanford parser
• GATE etc
• The issue is that this is expensive – in terms of detection time
• Due to models are complex or not optimised
• Tweet stream is too fast, …
• How do we reduce the cost of detecting entities?
• ….