ADAPTATION OF CYCLIC FRAMEWORK ON SENSOR NETWORK TO CALCULATE TRUSTWORTHINESS OF DATA ON WEB

• Main Author: Prismayadi , Wim
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/31635
• Institution: Institut Teknologi Bandung

With the fast growing of data on web, it is easier nowadays for people to search and retrieve the information they need. Therefore, the quality of data becomes an even moreimportant thing to be considered. Data quality can be evaluated by analyzing the trusthworthiness of the data, i.e. to see whether the data can be trusted or not, using data provenance. Provenance is defined as origin of data: where the data comes from and how it ends up at the current source of data. <br /> <br /> <br /> This thesis proposes a method to assess data trustworthiness based on provenance by adapting a cyclic framework for sensor network. This framework was used to assess data from sensor network by using value similarity and provenance similarity. This framework can only be used to calculate numerik type of data. Since there are different types of data on web, this thesis proposes the use of several data similarity measurement methods as well as an adaptation of the measurement of provenance based on the sources of web data. <br /> <br /> <br /> The adapted cyclic framework to measure the trustworthiness of data on web, including its provenance model, has been tested on several types of data: numerik data, text data, and object collection data.