Technical Reports 2007

10 Publikation(en)

Comparing Web Development Platforms Through the Eyes of Professional Developers

Will Hardy

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-11

tr-b-07-14.pdf

Objective comparisons of web development platforms are much needed by development teams. The Plat Forms experiment sought to provide these objective comparisons, and this survey was designed to provide the subjective context for the experiment’s results. The survey, run in June 2007 asked 268 web developers to directly compare their two best known platforms for 12 aspects of software quality. The results of this survey ideally show where the objective results of the Plat Forms research agree with the current opinions of web developers, and where they disagree. In addition to the 3 platforms measured in the Plat Forms research (Java, Perl, PHP), Ruby and Python developers were also involved. Many of the survey results agreed with the objective research, for example PHP was less dependent on tools and frameworks. Some results differed from the Plat Forms research, for example, PHP was rated poorly for security and robustness.

Streaming Databases for Audio and Video Theory and Praxis

Margarita Esponda

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-10

tr-b-07-15.pdf

This paper offers a review of technological developments regarding massive storage devices and the emergence of new streaming services for audio and video on-demand. While the capacity of storage devices has been increasing exponentially, the price of a stored bit has been falling at an even faster rate. Audio and video files can be stored on-line in personal computers and computer networks. Accessing this information requires new type of databases capable of handling special types of queries: access through annotations and metadata, access by similarity, and access by feature search. Databases with in-built streaming capabilities would be extremely useful for transmitting the information to the end user, while providing at the same time consistency checks, indexing, reporting, and querying features. They should also be fast and scalable. This paper reviews recent academic and industrial projects, and describes an implementation of a streaming video servlet based on the Oracle SQL database, its Binary Large Object storage data type, and the Java Database Connectivity interface.

Tracking honey bee dances from sparse optical flow fields

Tim Landgraf, Raul Rojas

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-06

tr-b-07-11.pdf

Tracking honey bee dances precisely is a hard task since the bees are highly flexible and the background highly distractive and cluttered. We propose methods that cluster optical flow features using a fast hough transform. A bounding box is shifted according to the movement of the underlying optical flow features. Errors made are corrected in a second step by evaluation of coherence of motion for future frames.

An Exact Mathematical Programming Approach to Multiple RNA Sequence-Structure Alignmen

Markus Bauer, Gunnar W. Klau, and Knut Reinert

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-03

tr-b-07-07.pdf

One of the main tasks in computational biology is the computation of alignments of genomic sequences to reveal their commonalities. In case of DNA or protein sequences, sequence information alone is usually sufficient to compute reliable alignments. RNA molecules, however, build spatial conformations---the secondary structure---that are more conserved than the actual sequence. Hence, computing reliable alignments of RNA molecules has to take into account the secondary structure. We present a novel framework for the computation of exact multiple sequence-structure alignments: We give a graph-theoretic representation of the sequence-structure alignment problem and phrase it as an integer linear program. We identify a class of constraints that make the problem easier to solve and relax the original integer linear program in a Lagrangian manner. Experiments on a recently published benchmark show that our algorithms has a comparable performance than more costly dynamic programming algorithms, and outperforms all other approaches in terms of solution quality with an increasing number of input sequences.

Accurate Multiple Sequence-Structure Alignment of RNA Sequences Using Combinatorial Optimization

Markus Bauer, Gunnar W. Klau, and Knut Reinert

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-03

tr-b-07-06.pdf

Background: The discovery of functional non-coding RNA sequences has led to an increasing interest in algorithms related to RNA analysis. Traditional sequence alignment algorithms, however, fail at computing reliable alignments of low-homology RNA sequences. The spatial conformation of RNA sequences largely determines their function, and therefore RNA alignment algorithms have to take structural information into account. Results: We present a graph-based representation for sequence-structure alignments, which we model as an integer linear program (ILP). We sketch how we compute an optimal or near-optimal solution to the ILP using methods from combinatorial optimization, and present results on a recently published benchmark set for RNA alignments. Conclusions: The implementation of our algorithm yields better alignments in terms of two published scores than the other programs that we tested: This is especially the case with an increasing number of input sequences. Our program LaRA is freely available from http://www.planet-lisa.net.

Adaptive Textzeilensegmentierung zur Lösung des Vice Versa-Problems

Marco Block, Marte Ramirez Ortegon, Manuel Siebeneicher, Alexander Seibert, Raul Rojas

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-01

tr-b-07-04.pdf

Plat_Forms 2007: The Web Development Platform Comparison — Evaluation and Results

Lutz Prechelt

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-01

tr-b-07-10.pdf

"Plat_Forms" is a competition in which top-class teams of three professional programmers competed to implement the same requirements for a web-based system within 30 hours, each team using a different technology platform (Java EE, PHP, or Perl). Plat_Forms intends to provide new insights into the real (rather than purported) pros, cons, and emergent properties of each platform. This report describes the evaluation of the solutions delivered by the participants and of the development process and presents the evaluation methodology as well as the results in great detail. It analyzes many aspects of each solution, both external (usability, functionality, reliability, robustness, etc.) and internal (size, structure, flexibility, modifiability, etc.). The many results we obtained cover a wide spectrum: First, there are results that many people would have called "obvious" or "well known", say, that Perl solutions tend to be more compact than Java solutions. Second, there are results that contradict conventional wisdom, say, that our PHP solutions appear in some (but not all) respects to be actually at least as secure as the others. Finally, one result makes a statement we have not seen discussed previously: The amount of variation between the teams tends to be smaller for PHP than for the other platforms in a whole variety of different respects.

Plat_Forms 2007 Task: PbT

Lutz Prechelt

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-01

tr-b-07-03.pdf

This document contains the requirements for the system to be built by the participants of the Plat_Forms 2007 contest. The system is called PbT (People by Temperament). It is to be written within 30 hours by a team of three people. For further details about the contest, please see www.plat-forms.org.

SITT - A Simple Robust Scaleinvariant Text Feature Detector

Marco Block, Marte Ramirez Ortegon, Alexander Seibert, Jan Kretzschmar, Raul Rojas

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-01

tr-b-07-02.pdf

In this paper we present SITT, a simple robust scaleinvariant text feature detector for document mosaicing. Digital image stitching has been studied for several decades. SIFT-Features in combination with RANSAC algorithm are established to produce good panoramas. The main problem of realtime text document stitching is the size of the feature set created by SIFT-Features. We introduce SITT-Features to solve this problem. Our experiments denote that for document images SITT-Features produce faster good results than SIFT-Features.

A Survey on Recognition of On-Line Handwritten Mathematical Notation

Ernesto Tapia, Raul Rojas

Takustr. 9, 14195 Berlin, Germany: Freie Universität Berlin, Institute of Computer Science | 2007-01

tr-b-07-01.pdf

This report describes recent advances in the area of the recognition of on-line handwritten mathematical notation. We describe architectures, symbol classification methods, and techniques for the structural analysis of mathematical expressions. We also survey applications specialized for mathematical notation.