Locality-sensitive Hash Functions Research Articles

Approximately Nearest Neighborhood Image Search Using Unsupervised Hashing via Homogeneous Kernels

We propose an approximation search algorithm that uses additive homogeneous kernel mapping to search for an image approximation based on kernelized locality-sensitive hashing. To address problems related to the unstable search accuracy of an unsupervised image hashing function and degradation of the search-time performance with increases in the number of hashing bits, we propose a method that combines additive explicit homogeneous kernel mapping and image feature histograms to construct a search algorithm based on a locality-sensitive hashing function. Moreover, to address the problem of semantic gaps caused by using image data that lack type information in semantic modeling, we describe an approximation searching algorithm based on the homogeneous kernel mapping of similarities between pairs of images and dissimilar constraint relationships. Our image search experiments confirmed that the proposed algorithm can construct a locality-sensitive hash function more accurately, thereby effectively improving the similarity search performance.

Query-aware locality-sensitive hashing scheme for $$l_p$$ norm

The problem of c-Approximate Nearest Neighbor (c-ANN) search in high-dimensional space is fundamentally important in many applications, such as image database and data mining. Locality-Sensitive Hashing (LSH) and its variants are the well-known indexing schemes to tackle the c-ANN search problem. Traditionally, LSH functions are constructed in a query-oblivious manner, in the sense that buckets are partitioned before any query arrives. However, objects closer to a query may be partitioned into different buckets, which is undesirable. Due to the use of query-oblivious bucket partition, the state-of-the-art LSH schemes for external memory, namely C2LSH and LSB-Forest, only work with approximation ratio of integer $$c \ge 2$$cź2. In this paper, we introduce a novel concept of query-aware bucket partition which uses a given query as the "anchor" for bucket partition. Accordingly, a query-aware LSH function under a specific $$l_p$$lp norm with $$p \in (0,2]$$pź(0,2] is a random projection coupled with query-aware bucket partition, which removes random shift required by traditional query-oblivious LSH functions. The query-aware bucket partitioning strategy can be easily implemented so that query performance is guaranteed. For each $$l_p$$lp norm $$(p \in (0,2])$$(pź(0,2]), based on the corresponding p-stable distribution, we propose a novel LSH scheme named query-aware LSH (QALSH) for c-ANN search over external memory. Our theoretical studies show that QALSH enjoys a guarantee on query quality. The use of query-aware LSH function enables QALSH to work with any approximation ratio $$c > 1$$c>1. In addition, we propose a heuristic variant named QALSH$$^+$$+ to improve the scalability of QALSH. Extensive experiments show that QALSH and QALSH$$^+$$+ outperform the state-of-the-art schemes, especially in high-dimensional space. Specifically, by using a ratio $$c < 2$$c<2, QALSH can achieve much better query quality.

Toward Efficient Multi-Keyword Fuzzy Search Over Encrypted Outsourced Data With Accuracy Improvement

Keyword-based search over encrypted outsourced data has become an important tool in the current cloud computing scenario. The majority of the existing techniques are focusing on multi-keyword exact match or single keyword fuzzy search. However, those existing techniques find less practical significance in real-world applications compared with the multi-keyword fuzzy search technique over encrypted data. The first attempt to construct such a multi-keyword fuzzy search scheme was reported by Wang et al. , who used locality-sensitive hashing functions and Bloom filtering to meet the goal of multi-keyword fuzzy search. Nevertheless, Wang’s scheme was only effective for a one letter mistake in keyword but was not effective for other common spelling mistakes. Moreover, Wang’s scheme was vulnerable to server out-of-order problems during the ranking process and did not consider the keyword weight. In this paper, based on Wang et al. ’s scheme, we propose an efficient multi-keyword fuzzy ranked search scheme based on Wang et al. ’s scheme that is able to address the aforementioned problems. First, we develop a new method of keyword transformation based on the uni-gram, which will simultaneously improve the accuracy and creates the ability to handle other spelling mistakes. In addition, keywords with the same root can be queried using the stemming algorithm. Furthermore, we consider the keyword weight when selecting an adequate matching file set. Experiments using real-world data show that our scheme is practically efficient and achieve high accuracy.

Employing Ontology-Alignment and Locality-Sensitive Hashing to Improve Attribute Interoperability in Federated eID Systems

Achieving interoperability, i.e. creating identity federations between different Electronic identities (eID) systems, has gained relevance throughout the past years. A serious problem of identity federations is the missing harmonization between various attribute providers (APs). In closed eID systems, ontologies allow a higher degree of automation in the process of aligning and aggregating attributes from different APs. This approach does not work for identity federations, as each eID system uses its own ontology to represent its attributes. Furthermore, providing attributes to intermediate entities required to align and aggregate attributes potentially violates privacy rules. To tackle these problems, we propose the use of combined ontology-alignment (OA) approaches and locality-sensitive hashing (LSH) functions. We assess existing implementations of these concepts defining and using criteria that are special for identity federations. Obtained results confirm that proper implementations of these concepts exist and that they can be used to achieve interoperability between eID systems on attribute level. A prototype is implemented showing that combining the two assessment winners (AlignAPI for ontology-alignment and Nilsimsa for LSH functions) achieves interoperability between eID systems. In addition, the improvement obtained in the alignment process by combining the two assessment winners does not impact negatively the privacy of the user’s data, since no clear-text data is exchanged in the alignment process.

Stratified locality-sensitive hashing for accelerated physiological time series retrieval.

We introduce stratified locality-sensitive hashing (SLSH) for retrieving similar physiological waveform time series. SLSH further accelerates the sublinear retrieval time obtained by the standard locality-sensitive hashing (LSH) method. The standard family of locality-sensitive hash functions is limited to provide only a single perspective on the data due to its one-to-one relationship to a distinct distance function for measuring similarity. SLSH incorporates multiple locality-sensitive hash families with various distance functions enabling it to examine the data with more diverse and refined perspectives. We provide the procedures of SLSH with locality-sensitive hash families for the l1 and the cosine distances, and compare its performance to the standard LSH on an arterial blood pressure time series data extracted from the physiological waveform repository of the MIMIC2 database. The time to retrieve five most similar waveforms by SLSH is 14 times faster than the linear search and 1.7 times faster than the standard LSH when we allow 5% decrease in accuracy as a trade-off.

Multi-Granularity Locality-Sensitive Bloom Filter

In many applications, such as homeland security, image processing, social network, and bioinformatics, it is often required to support an approximate membership query (AMQ) to answer a question like “is an (query) object q near to at least one of the objects in the given data set $\Omega$ ?” However, existing techniques for processing AMQs require a key parameter, i.e., the distance value, to be defined in advance for the query processing. In this paper, we propose a novel filter, called multi-granularity locality-sensitive Bloom filter (MLBF), which can process AMQs with multiple distance granularities. Specifically, the MLBF is composed of two Bloom filters (BF), one is called basic multi-granularity locality-sensitive BF (BMLBF), and the other is called multi-granularity verification BF (MVBF). The BMLBF is used to store the data objects. It adopts an alignable locality-sensitive hashing (LSH) function family to support multiple granularities. The MVBF is used to reduce the false positive rate of the MLBF. The false negative rate of the MLBF is reduced by applying AND-constructions followed by an OR-construction. In addition, based on the MLBF structure, we suggest a more space-effective variant, called the MLBF*, to further reduce space cost. Theoretical analyses for estimating false positive/negative rates of the MLBF/MLBF* are given. Experiments using synthetic and real data show that the theoretical estimates are quite accurate, and the MLBF/MLBF* technique can handle AMQs with low false positive and negative rates for multiple distance granularities.

Query-aware locality-sensitive hashing for approximate nearest neighbor search

Locality-Sensitive Hashing (LSH) and its variants are the well-known indexing schemes for the c -Approximate Nearest Neighbor ( c -ANN) search problem in high-dimensional Euclidean space. Traditionally, LSH functions are constructed in a query-oblivious manner in the sense that buckets are partitioned before any query arrives. However, objects closer to a query may be partitioned into different buckets, which is undesirable. Due to the use of query-oblivious bucket partition, the state-of-the-art LSH schemes for external memory, namely C2LSH and LSB-Forest, only work with approximation ratio of integer c ≥ 2. In this paper, we introduce a novel concept of query-aware bucket partition which uses a given query as the "anchor" for bucket partition. Accordingly, a query-aware LSH function is a random projection coupled with query-aware bucket partition, which removes random shift required by traditional query-oblivious LSH functions. Notably, query-aware bucket partition can be easily implemented so that query performance is guaranteed. We propose a novel query-aware LSH scheme named QALSH for c -ANN search over external memory. Our theoretical studies show that QALSH enjoys a guarantee on query quality. The use of query-aware LSH function enables QALSH to work with any approximation ratio c &gt; 1. Extensive experiments show that QALSH outperforms C2LSH and LSB-Forest, especially in high-dimensional space. Specifically, by using a ratio c &lt; 2, QALSH can achieve much better query quality.

Enhanced Locality Sensitive Clustering in High Dimensional Space

A dataset can be clustered by merging the bucket indices that come from the random projection of locality sensitive hashing functions. It should be noted that for this to work the merging interval must be calculated first. To improve the feasibility of large scale data clustering in high dimensional space we propose an enhanced Locality Sensitive Hashing Clustering Method. Firstly, multiple hashing functions are generated. Secondly, data points are projected to bucket indices. Thirdly, bucket indices are clustered to get class labels. Experimental results showed that on synthetic datasets this method achieves high accuracy at much improved cluster speeds. These attributes make it well suited to clustering data in high dimensional space.

Locality-Sensitive Bloom Filter for Approximate Membership Query

In many network applications, Bloom filters are used to support exact-matching membership query for their randomized space-efficient data structure with a small probability of false answers. In this paper, we extend the standard Bloom filter to Locality-Sensitive Bloom Filter (LSBF) to provide Approximate Membership Query (AMQ) service. We achieve this by replacing uniform and independent hash functions with locality-sensitive hash functions. Such replacement makes the storage in LSBF to be locality sensitive. Meanwhile, LSBF is space efficient and query responsive by employing the Bloom filter design. In the design of the LSBF structure, we propose a bit vector to reduce False Positives (FP). The bit vector can verify multiple attributes belonging to one member. We also use an active overflowed scheme to significantly decrease False Negatives (FN). Rigorous theoretical analysis (e.g., on FP, FN, and space overhead) shows that the design of LSBF is space compact and can provide accurate response to approximate membership queries. We have implemented LSBF in a real distributed system to perform extensive experiments using real-world traces. Experimental results show that LSBF, compared with a baseline approach and other state-of-the-art work in the literature (SmartStore and LSB-tree), takes less time to respond AMQ and consumes much less storage space.

Organization of multimedia data for conceptual search based on ontologies

Abstract Nowadays, there is a large volume of semantically annotated multimedia data available in the Semantic Web. These data have originated from several different sources, generating new issues about their storage and retrieval. In this scenario, simple ontologies are commonly used to define knowledge domains and classify data into concepts, establishing relations between them. Such conceptual relationship may be measured by a similarity function which allows the search to be performed by similarity in an indexing system. The contribution of this paper is to propose how to organize multimedia data using this conceptual classification in LSH (Locality Sensitive Hashing) functions, facilitating the conceptual search in distributed systems like P2P networks.

Hash Graph Based Semi-supervised Learning Method and Its Application in Image Segmentation

Graph based semi-supervised learning (GSL) method runs slowly because of the need of much time to construct a neighbor graph. This paper presents a hash graph based semi-supervised learning (HGSL) method, which can search neighbors by locality sensitive hashing function and efficiently reduce the time for GSL to construct a neighbor graph. Image segmentation experiments show that HGSL has an improvement of 0.47 % in average segmenting accuracy, and can greatly reduce the segmenting time, e.g., it takes about 28.5 % of the time for GSL to segment an image with size of 300 × 800.

Optimal Hash Functions for Approximate Matches on the $n$-Cube

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> One way to find near-matches in large datasets is to use hash functions. In recent years locality-sensitive hash functions for various metrics have been given; for the Hamming metric projecting onto <emphasis emphasistype="italic"><formula formulatype="inline"><tex Notation="TeX">$k$</tex></formula></emphasis> bits is simple hash function that performs well. In this paper, we investigate alternatives to projection. For various parameters hash functions given by complete decoding algorithms for error-correcting codes work better, and asymptotically random codes perform better than projection. </para>

Fast Similarity Search for Learned Metrics

We introduce a method that enables scalable similarity search for learned metrics. Given pairwise similarity and dissimilarity constraints between some examples, we learn a Mahalanobis distance function that captures the examples' underlying relationships well. To allow sublinear time similarity search under the learned metric, we show how to encode the learned metric parameterization into randomized locality-sensitive hash functions. We further formulate an indirect solution that enables metric learning and hashing for vector spaces whose high dimensionality makes it infeasible to learn an explicit transformation over the feature dimensions. We demonstrate the approach applied to a variety of image data sets, as well as a systems data set. The learned metrics improve accuracy relative to commonly used metric baselines, while our hashing construction enables efficient indexing with learned distances and very large databases.

Melody-based retrieval in audio collections

Mid-level representations are increasingly used to bridge the gap between high-level (semantic) and low-level audio representations. A mid-level representation that integrates melodic and rhythmic aspects of a music signal is introduced. The representation is formed by first performing multi-pitch detection on consecutive audio frames, and then searching for dominant melodic lines within the detected pitches. Beat-tracking is also performed to yield a beat-synchronous representation, independent of tempo variations. The representation is used as the basis for a melody-based audio retrieval system. An approximate nearest neighbor search algorithm is employed to compare sections of the mid-level representation of the query to indexed sections of mid-level representations of songs in a collection. Symbolic queries may also be used. Results are ranked according to the number of matched sections. A locality sensitive hashing function for cosine similarity is used for indexing, and the point location in equal balls algorithm for searching. Retrieval was tested on the cover song identification task, where the goal is to retrieve different interpretations of a song in a collection. Results on a collection of 2,400 songs are presented. [Work supported by ARRS.]