Reference no: EM132494303

COMP20008 Elements of Data Assignment help and solution, Processing Project - assessment writing service - University of Melbourne

Learning outcome 1: To gain practical experience in written communication skills for data science projects.
Learning outcome 2: To practice a selection of processing and exploratory analysis techniques through visu- alisation discussed in lectures and workshops.
Learning outcome 3: To practice crawling and scraping data from the Internet.
Learning outcome 4: To practice using widely used Python library for data processing and gain experience using library functions which may be unfamiliar and which require consultation of ad- ditional documentation from resources on the Web.

Your Parts

You are to perform a small data science project including some data processing and analysis using Python. Your responses to Parts 1-5 must be contained in a single .py file. Specifically, you have the following Parts:

Part 1

Produce a csv file containing the URL and headline of each the articles your crawler has found found. The CSV file should have two column headings url and headline and be called Part1.csv.

Note: You might want to start with a smaller website to test your crawling implementation with this site

Part 2
For each article found in Part 1,

a) extract the name of the first player mentioned in the article. You can find a list of player names as part of the tennis.json file provided. We will assume the article is written about that player (and only that player).

b) extract the first complete match score identified in the article. You will need to use regular expressions to accomplish this. We will assume this score relates to the first named player in the article.

Produce a csv file containing the URL, headline, first player mentioned and first complete match score of each the articles your crawler has found. The csv file should have four column headings url, headline, player and score and be called Part2.csv.

Note: Some articles may not contain a player name and/or a match score. These articles can be discarded.

Part 3

For each article used in Part 2, identify the absolute value of the game difference. E.g. a 6-2 6-2 score has a game difference of 8, while a 6-4 4-6 6-4 score has a game difference of 2. The value is referred to as the game difference

Produce a csv file containing the player name and average game difference for each player that at least one article has been written about. The csv file should have two column head- ings player and avg game difference and be called Part3.csv.

Part 4

Generate a suitable plot showing five players that articles are most frequently written about and the number of times an article is written about that player.
Save this plot as a png file called Part4.png

Part 5

Generate a suitable plot showing the average game difference for each player that at least one article has been written about and their win percentage. You can find a player's win percentage in the tennis.json file.
Save this plot as a png file called Part5.png

Part 6

Write a 3-4 page report to communicate the process and activities undertaken in the project, the analysis, and some limitations. Specifically, the report should contain the following infor- mation:
• A description of the crawling method and a brief summary the output for Part 1.

• A description of how you scraped data from each page, including any regular expressions used for Part 2 and a brief summary of the output.
• An analysis of the information shown in the two plots produced for Parts 4 & 5, in- cluding a brief summary of the data used. The plots are to be shown (included) along with your analysis.
• A discussion of the appropriateness of associating the first named player in the article with the first match score.
• At least one suggested method for how you could figure out from the contents of the ar- ticle whether the first named player won or lost the match being reported on.
• A discussion of what other information could be extracted from the articles to bet- ter understand player performance and a brief suggestion for how this could be done.

Attachment:- Elements of Data.rar