Tired of irrelevant search results? This paper introduces a novel approach to web-based information retrieval, focusing on context-driven search to deliver highly relevant results for both professional and non-professional users. The study presents a new conceptual paradigm where search is initiated from a user-selected text query within a document. The key innovation lies in leveraging the surrounding text as "context" to guide the search process, automating query augmentation and refinement. This context-driven process involves semantic keyword extraction and clustering to generate new, augmented queries that are submitted to various search engines. Finally, search results are semantically reranked using the initial context, ensuring relevance. The IntelliZap system implements this paradigm, demonstrating its effectiveness in offering even inexperienced users an advanced search tool. Experimental results confirm that using context significantly improves the quality and relevance of search results, offering a powerful alternative to traditional keyword-based search engines.
Published in ACM Transactions on Information Systems, this research aligns with the journal's focus on information retrieval and management systems. The proposed search paradigm directly addresses challenges in information retrieval by leveraging context to improve search relevance, a key topic in the field. The development and evaluation of the IntelliZap system demonstrate practical applications of information system principles, aligning with the journal's emphasis on both theoretical and practical contributions.