The government's efforts in developing electronic-based government services by utilizing information technology are referred to as the concept of e-government. Tangerang City is one of the cities that applies e-government in an application called Tangerang LIVE. In the Tangerang LIVE application, a LAKSA feature is used as a place for complaints from the people of Tangerang. This research was conducted to classify complaint data and determine the priority of groups of complaints received from the LAKSA feature. The technique used to conduct this research is clustering using the unsupervised learning method and the k-means algorithm, which will classify and predict the class for each document. In addition, an analysis of the priority complaint data was carried out based on the group that was received the most. The analysis carried out is to find out the class predictions for each complaint received, and then labeling will be given so that the complaint belongs to a more specific group. The results of the predictions will be displayed in the browser using web services.
TopIntroduction
Digital transformation is a new challenge for every country in the era of Industrial Revolution 4.0. Implementation of Smart City in Indonesia that creates the concept of smart governance be a challenge. Smart Governance is one of the dimensions of Smart Cities, which includes all aspects of political involvement and community service, as well as local government operations (Lopes, 2017). In implementing the smart city concept, the government is required to be able to provide governance that is transparent, accountable, collaborative (involving all stakeholders), and participatory (i.e., citizen participation) (Lopes, 2017). One example is providing a digital platform for the public to submit complaints to the government regarding the problems they are experiencing in their respective regions.
The Tangerang City Government has launched the Tangerang LIVE application. This application can help people to find information about Tangerang City. In the application, there is a feature to make a complaint called LAKSA or “Layanan Aspirasi Kotak Saran Anda”. The public can submit complaints about education, health, infrastructure, security, and social affairs in Tangerang City.
However, with the complaints, aspirations and complaints received, the Government must determine the complaints and complaints that must be followed up first. Therefore, from the amount of data received by the Government, it must be sorted according to the priority of complaints so that the problems with the most complaints will be immediately followed up by the stored parties. The amount of data on complaints and complaints received by the Government can be defined as Big Data. The term “big data” refers to data that is so large, fast or complex that it is difficult or impossible to use traditional methods. Big Data is a combination of structured, semi-structured, and unstructured data collected by an organization that can be used to obtain information and be used in machine learning projects, predictive modeling, and other analytical applications (Rouse, 2014).
The data used to perform the analysis is text, then the Data Mining method that will be used is Text Mining. Text Mining can also be referred to as Text Data Mining. Text mining refers to the process of extracting interesting and inappropriate patterns or knowledge from unstructured data. According to (Tan, 1999), Text Mining is believed to have more potential than Data Mining. Recent research has shown that 80% of company information is contained in text documents. Text Mining is a multidisciplinary field that involves information search, text analysis, grouping information, grouping, categorization, visualization, database technology, machine learning, and data mining. Text Mining aims to extract information from unstructured data and semi-structured data and find new patterns and information that is difficult to obtain without doing in-depth analysis.
With a large number of incoming complaint data, data processing will become more complicated and time-consuming. Therefore, an approach is needed to group the data to trigger follow-up or decision-making from the data.
The analytical method used is clustering, by identifying subgroups in the data so that the data points in the same subgroup (cluster) are very similar, while the data points in different groups are very different. Or it can be concluded to find homogeneous subgroups in the data so that the data points in each cluster are as similar as possible based on similarity measurements such as Euclidean-based distance or correlation-based distance. Each of the clusters formed will be a complaint category to assist the Tangerang City Government in dealing with problems.
Unlike supervised learning, the clustering method is considered an unsupervised learning method because there is no basic truth to compare the output of the clustering algorithm with the actual label to evaluate performance.