this section, we have presented the development of a Smart Context-aware Invoice Platform (SCIP) utilizing the Semantic-aware Service into Cloud Computing Architecture (SSCCA) proposed earlier. A detailed example has been provided to facilitate a better understanding of SSCCA and its potential applications.
The message flow shown in Fig. 4 can be explained through the following steps:
(1) The user completes a purchase using the e-invoice carrier provided by the Taiwan Ministry of Finance, and the purchase information is recorded on the user's cell phone using a one-dimensional barcode (refer to Fig. 5).
(2) The store transmits the purchase information to the user using NFC technology.
(3) The store uploads the purchase information to the Ministry of Finance's e-invoice database.
(4) The purchase information can be acquired using the e-invoice API, which is integrated into a PHP-based web platform.
(5) The PHP analyzes the JSON-formatted purchase information to extract the product name, which is then classified using an improved SVMV algorithm based on the Drink ontology classes.
(6) An RDF is generated based on the SVMV classification and the ontology, which represents the user's purchase preferences (refer to Fig. 7).
(7) The user performs exercise activities using context-awareness technology, with calorie consumption being calculated using the G-sensor (refer to Fig. 6).
(8) User context is acquired with the G-sensor. The information is converted to exercise status, which is used to calculate calorie consumption along with other related information.
(9) The calorie consumption information acquired in (8) will be stored in the cell phone database Android SQLite, which the user can access without internet connection.
(10) The context information, including calorie consumption, is uploaded to the web site database using PHP for processing.
(11) The PHP stores the information in the database and generates RDF based on the ontology.
(12) The Jena inference engine processes the RDF generated in steps 6 and 11 to produce a calorie intake recommendation based on the user's food purchases, calorie consumption, and other personal information.
(13) The semantic web inference results are displayed on the web site for the user to access.
(14) The context-awareness technology provides product suggestions based on the user's purchase behavior. Purchase history is acquired in step 4, and an RDF for products on sale is generated and processed on the Hadoop platform.
(15) Hadoop calculates the RDF from multiple users and makes purchase suggestions based on user preferences.
The presented research builds a Smart Context-aware Invoice Platform (SCIP) acting as a broker for Software as a Service (SaaS) utilizing the Semantic-aware Service into Cloud Computing Architecture (SSCCA) to gather personal electronic invoices for delivering context-aware services tailored to individual requirements. By integrating ontology technologies, the proposed SCIP platform demonstrates how a user profile based on Resource Description Framework (RDF) and Hadoop technologies can enhance personalized search. The SCIP interface is depicted in Figure 6.
Context-aware technology was utilized to gather contextual information from users, while sensors were also employed to collect data on human behaviors and inputs. The collected data was then processed through a server to provide users with relevant services. In this study, contextual information was acquired through context-aware technology, and users' preference information was generated in PHP using the resource description framework (RDF). The RDF generated from users' preference information can be utilized by Hadoop for computation, and the SCIP can display the results to recommend discounted products. SCIP generates RDF for information related to discounted products, and recommendations are made based on users' preferences. Furthermore, the platform transmits the RDF to the Hadoop Distributed File System (HDFS), which serves as Hadoop's data source. The RDF is then computed in Hadoop MapReduce, allowing for parallel computing on multi-user systems.