Filippo Geraci

STIMO: STIll and MOving video storyboard for the web scenario

In the current Web scenario a video browsing tool that produces on-the-fly storyboards is more and more a need. Video summary techniques can be helpful but, due to their long processing time, they are usually unsuitable for on-the-fly usage. Therefore, it is common to produce storyboards in advance, penalizing users customization. The lack of customization is more and more critical, as users have different demands and might access the Web with several different networking and device technologies. In this paper we propose STIMO, a summarization technique designed to produce on-the-fly video storyboards. STIMO produces still and moving storyboards and allows advanced users customization (e.g., users can select the storyboard length and the maximum time they are willing to wait to get the storyboard). STIMO is based on a fast clustering algorithm that selects the most representative video contents using HSV frame color distribution. Experimental results show that STIMO produces storyboards with good quality and in a time that makes on-the-fly usage possible.

Extraction and classification of dense communities in the web

The World Wide Web (WWW) is rapidly becoming important for society as a medium for sharing data, information and services, and there is a growing interest in tools for understanding collective behaviors and emerging phenomena in the WWW. In this paper we focus on the problem of searching and classifying communities in the web. Loosely speaking a community is a group of pages related to a common interest. More formally communities have been associated in the computer science literature with the existence of a locally dense sub-graph of the web-graph (where web pages are nodes and hyper-links are arcs of the web-graph). The core of our contribution is a new scalable algorithm for finding relatively dense subgraphs in massive graphs. We apply our algorithm on web-graphs built on three publicly available large crawls of the web (with raw sizes up to 120M nodes and 1G arcs). The effectiveness of our algorithm in finding dense subgraphs is demonstrated experimentally by embedding artificial communities in the web-graph and counting how many of these are blindly found. Effectiveness increases with the size and density of the communities: it is close to 100% for communities of a thirty nodes or more (even at low density). It is still about 80% even for communities of twenty nodes with density over 50% of the arcs present. At the lower extremes the algorithm catches 35% of dense communities made of ten nodes. We complete our Community Watch system by clustering the communities found in the web-graph into homogeneous groups by topic and labelling each group by representative keywords.

Cluster generation and cluster labelling for web snippets: a fast and accurate hierarchical solution

This paper describes Armil, a meta-search engine that groups into disjoint labelled clusters the Web snippets returned by auxiliary search engines. The cluster labels generated by Armil provide the user with a compact guide to assessing the relevance of each cluster to her information need. Striking the right balance between running time and cluster well-formedness was a key point in the design of our system. Both the clustering and the labelling tasks are performed on the fly by processing only the snippets provided by the auxiliary search engines, and use no external sources of knowledge. Clustering is performed by means of a fast version of the furthest-point-first algorithm for metric k-center clustering. Cluster labelling is achieved by combining intra-cluster and inter-cluster term extraction based on a variant of the information gain measure. We have tested the clustering effectiveness of Armil against Vivisimo, the de facto industrial standard in Web snippet clustering, using as benchmark a comprehensive set of snippets obtained from the Open Directory Project hierarchy. According to two widely accepted “external” metrics of clustering quality, Armil achieves better performance levels by 10%. We also report the results of a thorough user evaluation of both the clustering and the cluster labelling algorithms.

A comparison of several algorithms for the single individual SNP haplotyping reconstruction problem

Motivation: Single nucleotide polymorphisms are the most common form of variation in human DNA, and are involved in many research fields, from molecular biology to medical therapy. The technological opportunity to deal with long DNA sequences using shotgun sequencing has raised the problem of fragment recombination. In this regard, Single Individual Haplotyping (SIH) problem has received considerable attention over the past few years. Results: In this article, we survey seven recent approaches to the SIH problem and evaluate them extensively using real human haplotype data from the HapMap project. We also implemented a data generator tailored to the current shotgun sequencing technology that uses haplotypes from the HapMap project. Availability: The data we used to compare the algorithms are available on demand, since we think they represent an important benchmark that can be used to easily compare novel algorithmic ideas with the state of the art. Moreover, we had to re-implement six of the algorithms surveyed because the original code was not available to us. Five of these algorithms and the data generator used in this article endowed with a Web interface are available at http://bioalgo.iit.cnr.it/rehap Contact: filippo.geraci@iit.cnr.it

Extraction and classification of dense implicit communities in the Web graph

The World Wide Web (WWW) is rapidly becoming important for society as a medium for sharing data, information, and services, and there is a growing interest in tools for understanding collective behavior and emerging phenomena in the WWW. In this article we focus on the problem of searching and classifying communities in the Web. Loosely speaking a community is a group of pages related to a common interest. More formally, communities have been associated in the computer science literature with the existence of a locally dense subgraph of the Web graph (where Web pages are nodes and hyperlinks are arcs of the Web graph). The core of our contribution is a new scalable algorithm for finding relatively dense subgraphs in massive graphs. We apply our algorithm on Web graphs built on three publicly available large crawls of the Web (with raw sizes up to 120M nodes and 1G arcs). The effectiveness of our algorithm in finding dense subgraphs is demonstrated experimentally by embedding artificial communities in the Web graph and counting how many of these are blindly found. Effectiveness increases with the size and density of the communities: it is close to 100% for communities of thirty nodes or more (even at low density). It is still about 80% even for communities of twenty nodes with density over 50% of the arcs present. At the lower extremes the algorithm catches 35% of dense communities made of ten nodes. We also develop some sufficient conditions for the detection of a community under some local graph models and not-too-restrictive hypotheses. We complete our Community Watch system by clustering the communities found in the Web graph into homogeneous groups by topic and labeling each group by representative keywords.

SpeedHap: An Accurate Heuristic for the Single Individual SNP Haplotyping Problem with Many Gaps, High Reading Error Rate and Low Coverage

Single nucleotide polymorphism (SNP) is the most frequent form of DNA variation. The set of SNPs present in a chromosome (called the haplotype) is of interest in a wide area of applications in molecular biology and biomedicine, including diagnostic and medical therapy. In this paper we propose a new heuristic method for the problem of haplotype reconstruction for (portions of) a pair of homologous human chromosomes from a single individual (SIH). The problem is well known in literature and exact algorithms have been proposed for the case when no (or few) gaps are allowed in the input fragments. These algorithms, though exact and of polynomial complexity, are slow in practice. When gaps are considered no exact method of polynomial complexity is known. The problem is also hard to approximate with guarantees. Therefore fast heuristics have been proposed. In this paper we describe SpeedHap, a new heuristic method that is able to tackle the case of many gapped fragments and retains its effectiveness even when the input fragments have high rate of reading errors (up to 20%) and low coverage (as low as 3). We test SpeedHap on real data from the HapMap Project.

A scalable algorithm for high-quality clustering of web snippets

We consider the problem of partitioning, in a highly accurate and highly efficient way, a set of n documents lying in a metric space into k non-overlapping clusters. We augment the well-known furthest-point-first algorithm for k-center clustering in metric spaces with a filtering scheme based on the triangular inequality. We apply this algorithm to Web snippet clustering, comparing it against strong baselines consisting of recent, fast variants of the classical k-means iterative algorithm. Our main conclusion is that our method attains solutions of better or comparable accuracy, and does this within a fraction of the time required by the baselines. Our algorithm is thus valuable when, as in Web snippet clustering, either the real-time nature of the task or the large amount of data make the poorly scalable, traditional clustering methods unsuitable.

VISTO: visual storyboard for web video browsing

Web video browsing is rapidly becoming a very popular activity in the Web scenario, causing the production of a concise video content representation a real need. Currently, static video summary techniques can be used to this aim. Unfortunately, they require long processing time and hence all the summaries are produced in advance without any users customization. With an increasing number of videos and with the large users heterogeneousness, this is a burden. In this paper we propose VISTO, a summarization technique that produces customized on-the-fly video storyboards. The mechanism uses a fast clustering algorithm that selects the most representative frames using their HSV color distribution and allows users to select the storyboard length and the processing time. An objective and subjective evaluation shows that the storyboards are produced with good quality and in a time that allows on-the-fly usage.

K-Boost: a scalable algorithm for high-quality clustering of microarray gene expression data.

Microarray technology for profiling gene expression levels is a popular tool in modern biological research. Applications range from tissue classification to the detection of metabolic networks, from drug discovery to time-critical personalized medicine. Given the increase in size and complexity of the data sets produced, their analysis is becoming problematic in terms of time/quality trade-offs. Clustering genes with similar expression profiles is a key initial step for subsequent manipulations and the increasing volumes of data to be analyzed requires methods that are at the same time efficient (completing an analysis in minutes rather than hours) and effective (identifying significant clusters with high biological correlations). In this paper, we propose K-Boost, a clustering algorithm based on a combination of the furthest-point-first (FPF) heuristic for solving the metric k-center problem, a stability-based method for determining the number of clusters, and a k-means-like cluster refinement. K-Boost runs in O (|N| x k) time, where N is the input matrix and k is the number of proposed clusters. Experiments show that this low complexity is usually coupled with a very good quality of the computed clusterings, which we measure using both internal and external criteria. Supporting data can be found as online Supplementary Material at www.liebertonline.com.

A Fast and Accurate Heuristic for the Single Individual SNP Haplotyping Problem with Many Gaps, High Reading Error Rate and Low Coverage

Single nucleotide polymorphism (SNP) is the most frequent form of DNA variation. The set of SNPs present in a chromosome (called the haplotype) is of interest in a wide area of applications in molecular biology and biomedicine, including diagnostic and medical therapy. In this paper we propose a new heuristic method for the problem of haplotype reconstruction for (portions of ) a pair of homologous human chromosomes from a single individual (SIH). The problem is well known in literature and exact algorithms have been proposed for the case when no (or few) gaps are allowed in the input fragments. These algorithms, though exact and of polynomial complexity, are slow in practice. Therefore fast heuristics have been proposed. In this paper we describe a new heuristic method that is able to tackle the case of many gapped fragments and retains its effectiveness even when the input fragments have high rate of reading errors (up to 20%) and low coverage (as low as 3). We test our method on real data from the HapMap Project.