KEYWORD SEARCH IN TEXT CUBE: FINDING TOP-K RELEVANT CELLS BOLIN DING, YINTAO YU, BO ZHAO, CINDY XIDE LIN, JIAWEI HAN, AND CHENGXIANG ZHAI Abstract. We study the problem of keyword search in a data cube with text-rich dimension(s) (so-called text cube). The text cube is built on a multidimensional text database, where each row is associated with some text data (e.g., a document) and other structural dimensions (attributes). A cell in the text cube aggregates a set of documents with matching attribute values in a subset of dimensions. A cell document is the concatenation of all documents in a cell. Given a keyword query, our goal is to find the top-k most relevant cells (ranked according to the relevance scores of cell documents w.r.t. the given query) in the text cube. We define a keyword-based query language and apply IR-style relevance model for scoring and ranking cell documents in the text cube. We propose two efficient approaches to find the top-k answers. The proposed approaches support a general class of IR-style relevance scoring formulas that satisfy certain basic and common properties. One of them uses more time for pre-processing and less time for answering online queries; and the other one is more efficient in pre-processing and consumes more time for online queries. Experimental studies on the ASRS dataset are conducted to verify the efficiency and effectiveness of the proposed approaches.

Previous studies on supporting free-form keyword queries over RDBMSs provide users with linked-structures (e.g.,a set of joined tuples) that are relevant to a given keyword query. Most of them focus on ranking individual tuples from one table or joins of multiple tables containing a set of keywords. In this paper, we study the problem of keyword search in a data cube with text-rich dimension(s) (so-called text cube). The text cube is built on a multidimensional text database, where each row is associated with some text data (a document) and other structural dimensions (attributes). A cell in the text cube aggregates a set of documents with matching attribute values in a subset of dimensions. We define a keyword-based query language and an IR-style relevance model for coring/ranking cells in the text cube. Given a keyword query, our goal is to find the top-k most relevant cells. We propose four approaches, inverted-index one-scan, document sorted-scan, bottom-up dynamic programming, and search-space ordering. The search-space ordering algorithm explores only a small portion of the text cube for finding the top-k answers, and enables early termination. Extensive experimental studies are conducted to verify the effectiveness and efficiency of the proposed approaches. Citation: B. Ding, B. Zhao, C. X. Lin, J. Han, C. Zhai, A. N. Srivastava, and N. C. Oza, “Efficient Keyword-Based Search for Top-K Cells in Text Cube,” IEEE Transactions on Knowledge and Data Engineering, 2011.