Programming Assignment 4
Requirement: In this assignment, you are required to propose an interesting service mining problem and use current data mining/machine learning methods to solve the problem. You are required to use the service data retrieved from ProgrammableWeb and you are allowed to reuse any result or tool generated from PA3.
Steps:
1. Propose an interesting service mining problem, such as the ones related to service clustering, classification, and recommendation. The proposal should cover the general idea of the service mining problem, the motivation, the benefits, and how it is related to what we have learned in the class. You can also use the idea proposed in the papers cited in the lecture slides. 2. Describe the plan of how you will address the problem, including the algorithms/tools you will use and the evaluation plan. 3. Preprocess the web service data and run the selected algorithms using existing data mining/machine learning tools (such as R, Weka, RapidMiner, Matlab, and Apache Mahout), present and evaluate the evaluation result.
(Note: It is highly suggested that you get the feedbacks from the instructor on the first two steps before the submission.)
Required deliverables:
1. A document that includes the proposal, the design of the web service data analysis system, the development process, the screenshots of the testing scenarios, and the evaluation and presentation of the analysis result. 2. A readme file that describe how to install and test the web service data analysis system. 3. The entire project packages for the system including the data storage and data query. Include all the related files, such as source code files, script files, data files, configuration files, and etc. You can use any language for the development.
Submission Information: 1. Generate a zip file and name the files as $lastname_$firstname_PA4. No .rar file will be accepted. 2. Submit the zip files to the corresponding dropbox folders by May 5th.
Note: 1. You are allowed to discuss and learn in groups. However, you must design, develop, and submit the entire assignment by yourself. If there are two submissions having the exact same design, which is not very likely if working independently, the students will be asked to revise the design and resubmit the work. 2. Plagiarism checking will be performed to all the submission in this course.