Presents a review of 200 references in content-based image retrieval. The paper starts with discussing the working conditions of content-based retrieval: patterns of use, types of pictures, the role of semantics, and the sensory gap. Subsequent sections discuss computational steps for image retrieval systems. Step one of the review is image processing for retrieval sorted by color, texture, and local geometry. Features for retrieval are discussed next, sorted by: accumulative and global features, salient points, object and shape features, signs, and structural combinations thereof. Similarity of pictures and objects in pictures is reviewed for each of the feature types, in close connection to the types and means of feedback the user of the systems is capable of giving by interaction. We briefly discuss aspects of system engineering: databases, system architecture, and evaluation. In the concluding section, we present our view on: the driving force of the field, the heritage from computer vision, the influence on computer vision, the role of similarity and of interaction, the need for databases, the problem of evaluation, and the role of the semantic gap.

/pdf/content-based-image-retrieval-at-the-end-of-the-early-years-q45iy9ejai.pdf

Content-based image retrieval at the end of the early years

In this overview paper we motivate the need for and research issues arising from a new model of data processing. In this model, data does not take the form of persistent relations, but rather arrives in multiple, continuous, rapid, time-varying data streams. In addition to reviewing past work relevant to data stream systems and current projects in the area, the paper explores topics in stream query languages, new requirements and challenges in query processing, and algorithmic issues.

/pdf/models-and-issues-in-data-stream-systems-3aa6f5lsfv.pdf

Models and issues in data stream systems

The book Knowledge Discovery in Databases, edited by Piatetsky-Shapiro and Frawley [PSF91], is an early collection of research papers on knowledge discovery from data. The book Advances in Knowledge Discovery and Data Mining, edited by Fayyad, Piatetsky-Shapiro, Smyth, and Uthurusamy [FPSSe96], is a collection of later research results on knowledge discovery and data mining. There have been many data mining books published in recent years, including Predictive Data Mining by Weiss and Indurkhya [WI98], Data Mining Solutions: Methods and Tools for Solving Real-World Problems by Westphal and Blaxton [WB98], Mastering Data Mining: The Art and Science of Customer Relationship Management by Berry and Linofi [BL99], Building Data Mining Applications for CRM by Berson, Smith, and Thearling [BST99], Data Mining: Practical Machine Learning Tools and Techniques by Witten and Frank [WF05], Principles of Data Mining (Adaptive Computation and Machine Learning) by Hand, Mannila, and Smyth [HMS01], The Elements of Statistical Learning by Hastie, Tibshirani, and Friedman [HTF01], Data Mining: Introductory and Advanced Topics by Dunham, and Data Mining: Multimedia, Soft Computing, and Bioinformatics by Mitra and Acharya [MA03]. There are also books containing collections of papers on particular aspects of knowledge discovery, such as Machine Learning and Data Mining: Methods and Applications edited by Michalski, Brakto, and Kubat [MBK98], and Relational Data Mining edited by Dzeroski and Lavrac [De01], as well as many tutorial notes on data mining in major database, data mining and machine learning conferences.

/pdf/data-mining-concepts-and-techniques-2nd-edition-zx1xx1f5e3.pdf

Data Mining: Concepts and Techniques (2nd edition)

How do real graphs evolve over time? What are "normal" growth patterns in social, technological, and information networks? Many studies have discovered patterns in static graphs, identifying properties in a single snapshot of a large network, or in a very small number of snapshots; these include heavy tails for in- and out-degree distributions, communities, small-world phenomena, and others. However, given the lack of information about network evolution over long periods, it has been hard to convert these findings into statements about trends over time.Here we study a wide range of real graphs, and we observe some surprising phenomena. First, most of these graphs densify over time, with the number of edges growing super-linearly in the number of nodes. Second, the average distance between nodes often shrinks over time, in contrast to the conventional wisdom that such distance parameters should increase slowly as a function of the number of nodes (like O(log n) or O(log(log n)).Existing graph generation models do not exhibit these types of behavior, even at a qualitative level. We provide a new graph generator, based on a "forest fire" spreading process, that has a simple, intuitive justification, requires very few parameters (like the "flammability" of nodes), and produces graphs exhibiting the full range of properties observed both in prior work and in the present study.

/pdf/graphs-over-time-densification-laws-shrinking-diameters-and-59fgruyda7.pdf

Graphs over time: densification laws, shrinking diameters and possible explanations

(MATH) A locality sensitive hashing scheme is a distribution on a family $\F$ of hash functions operating on a collection of objects, such that for two objects x,y, PrheF[h(x) = h(y)] = sim(x,y), where sim(x,y) e [0,1] is some similarity function defined on the collection of objects. Such a scheme leads to a compact representation of objects so that similarity of objects can be estimated from their compact sketches, and also leads to efficient algorithms for approximate nearest neighbor search and clustering. Min-wise independent permutations provide an elegant construction of such a locality sensitive hashing scheme for a collection of subsets with the set similarity measure sim(A,B) = \frac{|A ∩ B|}{|A ∪ B|}.(MATH) We show that rounding algorithms for LPs and SDPs used in the context of approximation algorithms can be viewed as locality sensitive hashing schemes for several interesting collections of objects. Based on this insight, we construct new locality sensitive hashing schemes for:A collection of vectors with the distance between → \over u and → \over v measured by O(→ \over u, → \over v)/π, where O(→ \over u, → \over v) is the angle between → \over u) and → \over v). This yields a sketching scheme for estimating the cosine similarity measure between two vectors, as well as a simple alternative to minwise independent permutations for estimating set similarity.A collection of distributions on n points in a metric space, with distance between distributions measured by the Earth Mover Distance (EMD), (a popular distance measure in graphics and vision). Our hash functions map distributions to points in the metric space such that, for distributions P and Q, EMD(P,Q) ≤ Ehe\F [d(h(P),h(Q))] ≤ O(log n log log n). EMD(P, Q).

/pdf/similarity-estimation-techniques-from-rounding-algorithms-1i3hjl99zj.pdf

Similarity estimation techniques from rounding algorithms

Inherent in the operation of many decision support and continuous referral systems is the notion of the “influence” of a data point on the database. This notion arises in examples such as finding the set of customers affected by the opening of a new store outlet location, notifying the subset of subscribers to a digital library who will find a newly added document most relevant, etc. Standard approaches to determining the influence set of a data point involve range searching and nearest neighbor queries.In this paper, we formalize a novel notion of influence based on reverse neighbor queries and its variants. Since the nearest neighbor relation is not symmetric, the set of points that are closest to a query point (i.e., the nearest neighbors) differs from the set of points that have the query point as their nearest neighbor (called the reverse nearest neighbors). Influence sets based on reverse nearest neighbor (RNN) queries seem to capture the intuitive notion of influence from our motivating examples.We present a general approach for solving RNN queries and an efficient R-tree based method for large data sets, based on this approach. Although the RNN query appears to be natural, it has not been studied previously. RNN queries are of independent interest, and as such should be part of the suite of available queries for processing spatial and multimedia data. In our experiments with real geographical data, the proposed method appears to scale logarithmically, whereas straightforward sequential scan scales linearly. Our experimental study also shows that approaches based on range searching or nearest neighbors are ineffective at finding influence sets of our interest.

Influence sets based on reverse nearest neighbor queries

In many applications from telephone fraud detection to network management, data arrives in a stream, and there is a need to maintain a variety of statistical summary information about a large number of customers in an online fashion. At present, such applications maintain basic aggregates such as running extrema values (MIN, MAX), averages, standard deviations, etc., that can be computed over data streams with limited space in a straightforward way. However, many applications require knowledge of more complex aggregates relating different attributes, so-called correlated aggregates. As an example, one might be interested in computing the percentage of international phone calls that are longer than the average duration of a domestic phone call. Exact computation of this aggregate requires multiple passes over the data stream, which is infeasible.We propose single-pass techniques for approximate computation of correlated aggregates over both landmark and sliding window views of a data stream of tuples, using a very limited amount of space. We consider both the case where the independent aggregate (average duration in the example above) is an extrema value and the case where it is an average value, with any standard aggregate as the dependent aggregate; these can be used as building blocks for more sophisticated aggregates. We present an extensive experimental study based on some real and a wide variety of synthetic data sets to demonstrate the accuracy of our techniques. We show that this effectiveness is explained by the fact that our techniques exploit monotonicity and convergence properties of aggregates over data streams.

On computing correlated aggregates over continual data streams

Aggregation along hierarchies is a critical summary technique in a large variety of on-line applications including decision support and network management (e.g., IP clustering, denial-of-service attack monitoring). Despite the amount of recent study that has been dedicated to online aggregation on sets (e.g., quantiles, hot items), surprisingly little attention has been paid to summarizing hierarchical structure in stream data.

The problem we study in this paper is that of finding Hierarchical Heavy Hitters (HHH): given a hierarchy and a fraction φ, we want to find all HHH nodes that have a total number of descendants in the data stream no smaller than φ of the total number of elements in the data stream, after discounting the descendant nodes that are HHH nodes. The resulting summary gives a topological "cartogram" of the hierarchical data. We present deterministic and randomized algorithms for finding HHHs, which builds upon existing techniques by incorporating the hierarchy into the algorithms. Our experiments demonstrate several factors of improvement in accuracy over the straightforward approach, which is due to making algorithms hierarchy-aware.

/pdf/finding-hierarchical-heavy-hitters-in-data-streams-rf2qxqt8r4.pdf

Finding hierarchical heavy hitters in data streams

Spatial queries in high-dimensional spaces have been studied extensively. Among them, nearest neighbor queries are important in many settings, including spatial databases (Find the k closest cities) and multimedia databases (Find the k most similar images). Previous analyses have concluded that nearest-neighbor search is hopeless in high dimensions due to the notorious "curse of dimensionality". We show that this may be overpessimistic. We show that what determines the search performance (at least for R-tree-like structures) is the intrinsic dimensionality of the data set and not the dimensionality of the address space (referred to as the embedding dimensionality). The typical (and often implicit) assumption in many previous studies is that the data is uniformly distributed, with independence between attributes. However, real data sets overwhelmingly disobey these assumptions; rather, they typically are skewed and exhibit intrinsic ("fractal") dimensionalities that are much lower than their embedding dimension, e.g. due to subtle dependencies between attributes. We show how the Hausdorff and Correlation fractal dimensions of a data set can yield extremely accurate formulas that can predict the I/O performance to within one standard deviation on multiple real and synthetic data sets.

On the "dimensionality curse" and the "self-similarity blessing"

Conditional functional dependencies (CFDs) have recently been proposed as a useful integrity constraint to summarize data semantics and identify data inconsistencies. A CFD augments a functional dependency (FD) with a pattern tableau that defines the context (i.e., the subset of tuples) in which the underlying FD holds. While many aspects of CFDs have been studied, including static analysis and detecting and repairing violations, there has not been prior work on generating pattern tableaux, which is critical to realize the full potential of CFDs.This paper is the first to formally characterize a "good" pattern tableau, based on naturally desirable properties of support, confidence and parsimony. We show that the problem of generating an optimal tableau for a given FD is NP-complete but can be approximated in polynomial time via a greedy algorithm. For large data sets, we propose an "on-demand" algorithm providing the same approximation bound, that outperforms the basic greedy algorithm in running time by an order of magnitude. For ordered attributes, we propose the range tableau as a generalization of a pattern tableau, which can achieve even more parsimony. The effectiveness and efficiency of our techniques are experimentally demonstrated on real data.

/pdf/on-generating-near-optimal-tableaux-for-conditional-1obf5unf6k.pdf

Flip Korn

Papers

Influence sets based on reverse nearest neighbor queries

On computing correlated aggregates over continual data streams

Finding hierarchical heavy hitters in data streams

On the "dimensionality curse" and the "self-similarity blessing"

On generating near-optimal tableaux for conditional functional dependencies