The Open University Learning Analytics Dataset (OULAD) contains real data collected from
students in online courses at the Open University in the UK. The dataset includes information
about each student’s performance in the course and their interactions with online materials.
This dataset has become a benchmark for learning analytics researchers interested in
developing tools to support online learning, such as early warning systems to identify students
at risk of failing a course.
Like many datasets gathered from online systems, some of the data files are huge, making them
slow and cumbersome to analyze. For example, one file in the OULAD contains over 10.6
million rows and is so big it can’t be fully opened in standard spreadsheet software.
Your task is to write a program that will help automate analysis of the OULAD files. Download
the OULAD files and review the data description at the link above. The dataset contains more
files than you need for this assignment. This week, you will work with courses.csv (a small file)
and studentVle.csv (a huge file). You will need some of the other files in Assignment 6. All files
contain header rows.
The format of the data in courses.csv is:
code_module, code_presentation, code_presentation_length
• code_module . A module is equivalent to a course code, in Northeastern terms. For
example, CS5010 would be a course_module.
• code_presentation . A presentation is equivalent to an offering of a course. For
example, Fall2019 would be a course_presentation.
• code_presentation_length is the length of the course offering in days. (courses at
the Open University last a lot longer than courses at Northeastern).
The format of the data in studentVle is:
code_module, code_presentation, id_student, id_site, date, sum_click
• code_module and code_presentation are the same as in courses.csv.
studentVle.csv only contains data from courses included in courses.csv.
• id_student is the unique ID number of a student in the given course.
• id_site is the unique ID number of a particular online resource.
• date is a specific day within the course, relative to the course start date. For example,
-10 means 10 days before the course started.
• sum_click is the number of times the given student clicked on the given resource on
the given day.
For both parts below, you will need to write a program that will read in these files and use them
to create new summary files so that each individual code_presentation of a code_module
has a file listing the total number of clicks that occurred on a particular day. Each new file
should be named code_presentation_code_module.csv and should contain two columns: date,
For example, one code_module is “AAA”. It has two code_presentations, one of which is
“2013J” and the other is “2014J”. Therefore, your program should produce two csv files for this
module: AAA_2013J.csv and AAA_2014J.csv. Within each file, each row should contain one
date and the total number of clicks that occurred in the presentation across all id_sites and
all id_students. For example, one row in “AAA_2013J.csv” should contain -10, 11952,
meaning that there were 11952 clicks in this course presentation on day -10.
Part 1: Sequential solution
Create a package called sequentialSolution and write a sequential solution to the above
problem. Your main class should accept the directory containing the OULAD csv files as a
command line argument. You may hardcode the names of the two OULAD csv files: courses.csv
For testing, you will probably find it helpful to create a smaller version of studentVle.csv!
Part 2: Concurrent solution
Create a package called concurrentSolution. Redesign your solution to part 1 to use
threads in a “producer-consumer” style. The inputs and outputs will be the same but your code
now contains multiple threads.
For example, when you read from studentVle.csv, you can think of this as a producer that
simply knows how to read this given file type and write each line “somewhere” for another
thread, the consumer, to process. Consumers can read the contents and transform them into the
results required, and write these “somewhere” for maybe another thread to write the output.
Your solution must have producer and consumer threads running concurrently, and
communicating via some suitable shared data structure. Producers and consumers should be
loosely coupled and communicate only through some queue or channel.
Part 3: Identify high activity days
Modify your concurrent solution to take an additional integer value on the command line,
which we’ll call an activity “threshold”. The threshold is a value that the user wants to use to
identify days when each course had the most activity.
You should use the threshold as follows:
1. Process the summary files produced above to find each day in each file where the
total_clicks is >= the threshold value.
2. Create a new file called activity-threshold.csv, where “threshold” is the value entered by
the user that contains all rows identified in step 1. Your file will also need to indicate
which module and presentation the data came from, so it should contain the following
module_presentation, date, total_clicks
So, for example, if the threshold entered by the user was 11000, the file created would be
called activity-11000.csv and one row in the file would contain the following data:
AAA_2013J, -10, 11952, where -10 is the date and 11952 is the total number of clicks on
本网站支持淘宝 支付宝 微信支付 paypal等等交易。如果不放心可以用淘宝交易！
E-mail: [email protected] 微信:itcsdx