OFFICE DOCUMENT SEARCH ENGINE

• Main Author: Ardian Wirasandi, Diki
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39623
• Institution: Institut Teknologi Bandung

In the current digital era, electronic document has become a common occurrence especially in office jobs. However, finding a specific document becomes increasingly difficult as the number of the stored document also increases. A search engine that is able to recognize, store, and perform queries for various document formats would be needed to solve that problem. How to recognize the office document structure or pattern as well as how to build an office document search engine with optimal query result are the main focus of this thesis. Document structure or pattern recognition is done by performing word extraction and utilizing the words position in the document itself. The office document search engine system that has been developed consists of two main processes, which are indexing process and query process. In indexing process, the documents are gathered and processed into indexes which will be stored by the system. Query process consists of receiving user’s search query, processing the query, and generating document ranks based on the search results as well as displaying them. Evaluation result shows the system that has been developed performs well. Furthermore, every functional requirement and non-functional requirement have been fulfilled. This shows the method used to recognize the document structure or pattern is suitable for office documents processing. Based on the analysis of the evaluation result, it can be concluded that the method and workflow applied in the search engine system are able to produce optimal search result.