Differences

This shows you the differences between two versions of the page.

--- teaching:projh402 [2013/09/23 11:47]
svsummer
+++ teaching:projh402 [2016/10/24 15:32]
svsummer
@@ Line 7: / Line 7: @@
 ===== Project proposals =====
-==== Development of a Personal Scientific Digital Library Management System ====
+=== Fast loading of semantic web datasets into native RDF stores ====
-In this project, the student is asked to construct a software system to help manage large collections of scientific papers in digital form. Specifically, the system must be able to:
+The next generation of the Web, the so-called Semantic Web, stores
-  - Scan a given filesystem location for given filetypes (PDFs, EPUB, ...) containing scientific articles.
+extremely large knowledge bases in the RDF data model. In this data
-  - Extract the metadata from each identified file. Here, the metadata includes the title of the article, its authors, the publishing venue, the publisher, the year of publication, the article's abstract ... The development of an intelligent way to retreive this metadata is requried. This could be done, for example by a combination of parsing the file, contacting the internet repositories of known publishers (AMC, Springer, Elsevier) etc to retrieve the data.
+model, all knowledge is represented by means of triples of the form
-  - Offer search capabilities, in order to allow a user to find all indexed articles matching certain criteria (title, author, ...)
+(subject, property, object), where subject, property and object can be
-  - Offer archiving capabilities
+URLs, among other things.
+In order to effeciently query such knowledge bases, the RDF data is
+typically loaded into a so-called native RDF store. To ensure that the
+knowledge is encoded for fast retrieval, the RDF store will first
+encode all variable-length URLs in the dataset by fixed-width
+integers, among other things. Each RDF triple will then be encoded by
+by their corresponding integer triples (integer_of_subject,
+integer_of_property, integer_of_object).
+The purpose of this project is to implement and experementally compare
+a number of algorithms that can perform this encoding:
+  * The trivial alogrithm that simply maintains a hashmap that maps URLs to their integer codes. When processing triple (s,p,o), it looks up s, p, and o in the hashmap to see if they have alraedy been assigned an integer ID. If so, this id is used for the   encoding; otherwise they are inserted into the hashmap with new, unique ids. The downside of this approach is that, while simple, it requires that one can store all URLS in working memory.
+  * The slightly smarter algorithm that works in multiple stages: the ID is computed by a pre-fixed hash function. For each URL, the URL and its ID are written to an output file. This file is later sorted on ID to check for possible  hash collisions between distinct URLS.
+  * Algorithms that use the best known state-of-the art data structures for compactly storing respresenting sets of strings, such as the HAT-TRIE ("Engineering scalable, cache and space efficient tries for strings" Nikolas Askitis, Ranjan Sinha, The VLDB Journal, October 2010, Volume 19, Issue 5, pp 633-660 and "HAT-Trie: A Cache-Conscious Trie-Based Data Structure For Strings", The 30th International Australasian Computer Science Conference (ACSC), Volume 62, pages 97 - 105, 2007.).
+  * Variations of the above algorithms, fine-tuned for semantic web datasets.
-Use of semantic web technologies (RDF, SPARQL, ...) to store and search the metadata is encouraged.
 **Contact** : Stijn Vansummeren (stijn.vansummeren@ulb.ac.be)
 **Status**: available
+===== Graph Indexing for Fast Subgraph Isomorphism Testing =====
-==== Curriculum Revision Assistant ====
+There is an increasing amount of scientific data, mostly from the bio-medical sciences, that can be represented as collections of graphs (chemical molecules, gene interaction networks, ...). A crucial operation when searching in this data is that of subgraph    isomorphism testing: given a pattern P that one is interested in (also a graph) in and a collection D of graphs (e.g., chemical molecules), find all graphs in G that have P as a   subgraph. Unfortunately, the subgraph isomorphism problem is computationally intractable. In ongoing research, to enable tractable processing of this problem, we aim to reduce the number of candidate graphs in D to which a subgraph isomorphism test needs   to be executed. Specifically, we index the graphs in the collection D by means of decomposing them into graphs for which subgraph   isomorphism *is* tractable. An associated algorithm that filters graphs that certainly cannot match P can then formulated based on ideas from information retrieval.
-In this project, the student is asked to construct a software system that can assist in the revision of teaching curricula (also known as teaching programs). The system should have the following functionalities:
+In this project, the student will emperically validate on real-world datasets the extent to which graphs can be decomposed into graphs for which subgraph isomorphism is tractable, and run experiments to validate the effectiveness of the proposed method in terms of filtering power.
-  - It should be able to load existing curricula from the ULB central administration. This could be done, for example, by parsing the webpages available at banner (the Civil Engineering in CS program is available at http://banssbfr.ulb.ac.be/PROD_frFR/bzscrse.p_disp_prog_detail?term_in=201314&prog_in=MA-IRIF&lang=FRENCH, for example).
-  - It should allow to make different versions of the teaching programs, much in the same way as version control systems like GIT and subversion offer the possibility to make different "development branches" of a program's source code.
+**Interested?** Contact : [[stijn.vansummeren@ulb.ac.be|Stijn Vansummeren]]
-  - It should allow to analyze the modifications proposed in the teaching programs, and summarize the impact that these changes could have on other programs. (For example, if a course is removed from the computer science curriculum, it should also be removed from all curricula that included the course.)
+**Status**: available
+==== Principles of Database Management Architectures in Managed Virtual Environments ====
+With the gaining popularity of Big Data, many data processing engines
+are implemented in a managed virtual environment such as the Java
+Virtual Machine (e.g., Apache Hadoop, Apache Giraph, Drill,
+...). While this improves the portability of the engine, the tradeoffs
+and implementation principles w.r.t. traditional C++ implementations
+are sometimes less understood.
+The objective in this project is to develop some basic functionalities
+of a database storage engine (Linked files, BTree, Extensible Hash
+table, basic external-memory sorting ) in a managed virtual machine
+(i.e., the Java Virtual Machine or and the .NET Common Language
+Runtime), and compare this with a C++-based implementation both on (1)
+ease of implementation and (2) execution efficiency. In order to
+develop the managed virtual machine implementation, the interested
+student will need to research the best practices that are used in the
+above-mentioned projects to gain maximum execution speed (e.g., use of
+the java.lang.unsafe feature, memory-mapped files, ...).
+**Contact** : Stijn Vansummeren (stijn.vansummeren@ulb.ac.be)
+**Status**: taken.
+==== Development of a distributed simulation algorithm ====
+Simulation and Bisimulation are fundamental notions in computer
+science. They underly many formal verification algorithms, and have
+recently been applied to the construction of so-called structural
+indexes,which are novel index data structures for relational databases
+and the Semantic Web.  Essentially, a (bi)simulation is a relation on
+the nodes of a graph. Unfortunately, however, while efficient
+main-memory algorithms for computing whether two nodes are similar
+exist, these algorithms fail when no the input graphs are too large to
+fit in main memory.
+The objective of this project is to implement a recently proposed
+algorithm for  computing simulation in a distributed setting, and
+provide a preliminary performance evaluation of this implementation.
 **Contact** : Stijn Vansummeren (stijn.vansummeren@ulb.ac.be)
 **Status**: available