The Implementation of Preference Ranking-based Spam Filtering in the Design of Company Complaint Reporting System

Abstract

A company complaint system is an important tool for identifying and resolving workplace problems. By implementing transparent, supportive and responsive mechanisms, companies can create a positive work culture and empower employees to report concerns and contribute to overall organizational improvement. However, in the company's complaint system, there are still spam complaints so the process of receiving complaints in a system becomes difficult. The methodology used for research related to spam complaints is TF-IDF Preference Ranking, which involves determining the order of spam and non-spam complaint categories based on similarity values to determine complaint categories based on weighted documents. Blackbox testing of the research results shows that TF-IDF Preference Ranking can filter spam effectively by comparing complaint data with the system's spam dataset. The test results show that 90% of the spam test data can be categorized as spam and 95% of the functional requirements have passed the functional test and are in accordance with user needs.