Finding and approximating top-k answers in keyword proximity search

Abstract

Various approaches for keyword proximity search have been implemented in relational databases, XML and the Web. Yet, in all of them, an answer is a Q-fragment, namely, a subtree T of the given data graph G, such that T contains all the keywords of the query Q and has no proper subtree with this property. The rank of an answer is inversely proportional to its weight. Three problems are of interest: finding an optimal (i.e., top-ranked) answer, computing the top-k answers and enumerating all the answers in ranked order. It is shown that, under data complexity, an efficient algorithm for solving the first problem is sufficient for solving the other two problems with polynomial delay. Similarly, an efficient algorithm for finding a θ-approximation of the optimal answer suffices for carrying out the following two tasks with polynomial delay, under query-and-data complexity. First, enumerating in a (θ + 1)-approximate order. Second, computing a (θ + 1)-approximation of the top-k answers. As a corollary, this paper gives the first efficient algorithms, under data complexity, for enumerating all the answers in ranked order and for computing the top-k answers. It also gives the first efficient algorithms, under query-and-data complexity, for enumerating in a provably approximate order and for computing an approximation of the top-k answers.

Learning to create data-integrating queries

“…In addition, in the previous work [27], the graph represents actual data items and their associations, and the Steiner trees are possible answers containing given keywords. Since data graphs can be very large, the method is primarily of theoretical rather than practical interest.…”

Processing Keyword Queries Under Access Limitations

“…In addition, in [6] it is shown that the optimal q-fragments of r can be enumerated in ranked-order with polynomial delay, i.e., the time for printing the next optimal answer is again polynomial in the size of r. Hence, we can state the following preliminary result. Theorem 1.…”

“…An efficient method for solving this problem in the context of keyword search over structured data is presented in [6], where a q-fragment can model our notion of answer. Yet, when optimality is not required, a simple technique (quadratic in the size of r) to obtain an answer (steps 2-6 of Algorithm 3) consists in trying to remove any tuple from the set as long as it contains all the keywords and remains connected.…”

Structured search result differentiation

“…We first compute s 1,l for each l (lines 2-5), then compute s k,l as discussed. Meanwhile, we record array best, which is used to re-produce the optimal DFS, newDF S (lines [11][12][13][14][15][16][17]. Finally, DoD is calculated by comparing newDF S with every other DFS (lines 18-21).…”

“…Many approaches have been proposed for supporting keyword search on XML data [4,8,9,14,15,16,17,19,22,24,18], and keyword search on graphs / relational databases [10,23,1,2,12,20,6,16,13,7]. Various ranking schemes have been proposed in these studies, including IR-style ranking (term frequency, document frequency, etc.…”