Weighted Mean of a Pair of Graphs

G and G′, with d(G, G′) being the edit distance of G and G′, the weighted mean of G and G′ is a graph G″ that has edit distances d(G, G″) and d(G″, G′) to G and G′, respectively, such that d(G, G″) + d(G″, G′) = d(G,G′). We'll show formal properties of the weighted mean, describe a procedure for its computation, and give examples. Received April 9, 2001     

SEPIA: estimating selectivities of approximate string predicates in large Databases

Many database applications have the emerging need to support approximate queries that ask for strings that are similar to a given string, such as “name similar to smith” and “telephone number similar to 412-0964”. Query optimization needs the selectivity of such an approximate predicate, i.e., the fraction of records in the database that satisfy the condition. In this paper, we study the problem of estimating selectivities of approximate string predicates. We develop a novel technique, called Sepia, to solve the problem. Given a bag of strings, our technique groups the strings into clusters, builds a histogram structure for each cluster, and constructs a global histogram. It is based on the following intuition: given a query string q, a preselected string p in a cluster, and a string s in the cluster, based on the proximity between q and p, and the proximity between p and s, we can obtain a probability distribution from a global histogram about the similarity between q and s. We give a full specification of the technique using the edit distance metric. We study challenges in adopting this technique, including how to construct the histogram structures, how to use them to do selectivity estimation, and how to alleviate the effect of non-uniform errors in the estimation. We discuss how to extend the techniques to other similarity functions. Our extensive experiments on real data sets show that this technique can accurately estimate selectivities of approximate string predicates. A short version of this article appeared as [21] in the proceedings of the 31st International Conference on Very Large Data Bases (VLDB), August 30 – September 2, 2005, Trondheim, Norway. The source code of our algorithms is available at .     

Approximating Tree Edit Distance through String Edit Distance

We present an algorithm to approximate edit distance between two ordered and rooted trees of bounded degree. In this algorithm, each input tree is transformed into a string by computing the Euler string, where labels of some edges in the input trees are modified so that structures of small subtrees are reflected to the labels. We show that the edit distance between trees is at least 1/6 and at most O(n ^3/4) of the edit distance between the transformed strings, where n is the maximum size of two input trees and we assume unit cost edit operations for both trees and strings. The algorithm works in O(n ²) time since computation of edit distance and reconstruction of tree mapping from string alignment takes O(n ²) time though transformation itself can be done in O(n) time.     

Edit distance for a run-length-encoded string and an uncompressed string

We propose a new algorithm for computing the edit distance of an uncompressed string against a run-length-encoded string. For an uncompressed string of length n and a compressed string with M runs, the algorithm computes their edit distance in time O(Mn). This result directly implies an O(min{mN,Mn}) time algorithm for strings of lengths m and n with M and N runs, respectively. It improves the previous best known time bound O(mN+Mn).     

A Fully Compressed Algorithm for Computing the Edit Distance of Run-Length Encoded Strings

A recent trend in stringology explores the possibility of utilizing text compression to speed up similarity computation between strings. In this line of investigation, run-length encoding is one of the earliest studied compression schemes. Despite its simple coding nature, the only positive result before this work is the computation of the in-del distance (dual of longest common subsequence), which requires O(mnlogmn) time, where m and n denote the number of runs of the input strings. The worst-case time complexity of computing the edit distance between two run-length encoded strings still depends on the uncompressed string lengths. In this paper, we break the foundational gap by providing its first “fully compressed” algorithm whose running time depends solely on the compressed string lengths. Specifically, given two strings, compressed into m and n runs, m≤n, we present an O(mn ²)-time algorithm for computing the edit distance of the strings. Our approach also yields the first fully compressed solution to approximate matching of a pattern of m runs in a text of n runs in O(mn ²) time.     

LCS approximation via embedding into locally non-repetitive strings

A classical measure of similarity between strings is the length of the longest common subsequence (LCS) between the two given strings. The search for efficient algorithms for finding the LCS has been going on for more than three decades. To date, all known algorithms may take quadratic time (shaved by logarithmic factors) to find large LCS. In this paper, the problem of approximating LCS is studied, while focusing on the hard inputs for this problem, namely, approximating LCS of near-linear size in strings over a relatively large alphabet (of size at least n^? for some constant ?>0, where n is the length of the string). We show that, any given string over a relatively large alphabet can be embedded into a locally non-repetitive string. This embedding has a negligible additive distortion for strings that are not too dissimilar in terms of the edit distance. We also show that LCS can be efficiently approximated in locally-non-repetitive strings. Our new method (the embedding together with the approximation algorithm) gives a strictly sub-quadratic time algorithm (i.e., of complexity O(n^2-?) for some constant ?) which can find common subsequences of linear (and near linear) size that cannot be detected efficiently by the existing tools.     

Approximate string matching with suffix automata

Theapproximate string matching problem is, given a text string, a pattern string, and an integerk, to find in the text all approximate occurrences of the pattern. An approximate occurrence means a substring of the text with edit distance at mostk from the pattern. We give a newO(kn) algorithm for this problem, wheren is the length of the text. The algorithm is based on the suffix automaton with failure transitions and on the diagonalwise monotonicity of the edit distance table. Some experiments showing that the algorithm has a small overhead are reported.     

Learning state machine-based string edit kernels

During the past few years, several works have been done to derive string kernels from probability distributions. For instance, the Fisher kernel uses a generative model M (e.g. a hidden Markov model) and compares two strings according to how they are generated by M. On the other hand, the marginalized kernels allow the computation of the joint similarity between two instances by summing conditional probabilities. In this paper, we adapt this approach to edit distance-based conditional distributions and we present a way to learn a new string edit kernel. We show that the practical computation of such a kernel between two strings x and x^′ built from an alphabet Σ requires (i) to learn edit probabilities in the form of the parameters of a stochastic state machine and (ii) to calculate an infinite sum over Σ^* by resorting to the intersection of probabilistic automata as done for rational kernels. We show on a handwritten character recognition task that our new kernel outperforms not only the state of the art string kernels and string edit kernels but also the standard edit distance used by a neighborhood-based classifier.     

Approximate string matching using withinword parallelism

Given a text string, a pattern string, and an integer k, the problem of approximate string matching with k differences is to find all substrings of the text string whose edit distance from the pattern string is less than k. The edit distance between two strings is defined as the minimum number of differences, where a difference can be a substitution, insertion, or deletion of a single character. An implementation of the dynamic programming algorithm for this problem is given that packs several characters and mod-4 integers into a computer word. Thus, it is a parallelization of the conventional implementation that runs on ordinary processors. Since a small alphabet means that characters have short binary codes, the degree of parallelism is greatest for small alphabets and for processors with long words. For an alphabet of size 8 or smaller and a 64 bit processor, a 21-fold parallelism over the conventional algorithm can be obtained. Empirical comparisons to the basic dynamic programming algorithm, to a version of Ukkonen's algorithm, to the algorithm of Galil and Park, and to a limited implementation of the Wu-Manber algorithm are given.     

支持块编辑距离的索引结构

在近似字符串匹配中,传统的编辑距离不能很好地衡量诸如人名、地址等数据的相似关系,而块编辑距离可以很好地衡量两个字符串的相似性.如何有效地支持块编辑距离,进行近似字符串查询处理具有重要的意义.计算两个字符串的块编辑距离是一个NP完全问题,因此希望提供有效的方法可以增强过滤能力,并减少假通过率.设计了一种支持移动编辑距离的新颖的索引结构SHV-Trie,通过研究移动编辑距离的操作特性,使用字母出现的频率作为支持移动编辑距离操作的一个下界,并且提出相应的查询过滤算法,同时,针对索引SHV-Trie的空间开销过大的问题,提出一种优化字母排列的索引结构和一种压缩的索引结构及相关查询过滤算法.真实数据集上的实验结果与分析显示了所提出的索引结构具有良好的过滤能力,并通过减少效率假通过率提高查询的效率.