A prototype for management of data in Open Source forges

Submitted by qadmin on Wed, 12/01/2010 - 17:28
in
Participate in our forum about information management in forges

Description

Forges are usually plenty of information, not always well-connected and stored. Thus, the search of information in these unstructured sources becomes a challenge for communities administrators and even for the developers or users. The result achieved and demostrated through this prototype mitigates the current situation, promising more efficient searching of forges. The prototype includes three different but complementary tools: conceptual search, semantic search and semantic navigator.

Conceptual search allows users query the information stored not by single terms, but by selecting a specific concept of the terms entered. Users should use this tool when not knowing the exact term that will bring up the information is looking for, but knowing the concept. Two types of search are available:

  • Simple search. After writing on the form (i.e. java) it can be automatically seen a list of possible unaware but related concepts for the user (an Island, a beverage, or a platform)
  • Advanced search and suggestions. Selecting a concept we can see the list of related results. Besides It is possible to filter results by a specific title, date, author and commitment.

Semantic Search is the frontend utility that takes advantage of the extracted information and allows users to perform queries searching for very specific results. Using this tool, users can select what kind of information are looking for (issues, sources, documents, etc.), which fields of the selected kind of information wants to be shown at the result page, and which are the conditions that the selected item must meet. The result page will show the selected fields among with a link to the original source of information. The result is also integrated with the semantic navigator, establishing a stating point for a navigation. This tool should be used when trying to find information that should meet certain known constraints.

Semantic Navigator is a tool that allows users viewing what is on the semantic repository and become aware of unknown relationships following the semantic links between entities. To start a navigation, it is necessary to establish a starting point. This can be done using the semantic search, performing a query using an embedded simple search, or selecting a known entity from the whole list of objects. The tool will show a representation of the information stored at the semantic repository regarding the starting entity: the properties of the object and its relation with other entities. The user can expand the hyperbolic tree through the links to discover more related semantic data. The semantic navigator tool is recommended when not knowing exactly what is looking for, but knowing that should have something to do with certain object, or when wishing to watch all semantic data regarding an object.

Download factsheet about information management for deeper understanding

Pre-conditions and requirements

These are the hardware and software requirements for the server repository:

  • Server machine able to run Java 1.4.2 or higher including 2 GB RAM, 10GB free disk and internet connection
  • Java virtual machine version 1.4.2 or higher
  • J2EE application server (Tomcat 4.1 for example)
  • MySQL database server 5.1

Associated documentation

Download

The prototype consist of 3 types of tools that you can find described here. Below the direct links to the online demonstration of them.

If you wish to donwload the source code for local installation and configuration, please follow the instructions at Documentation section and download herethe code.

Learned lessons

Conceptual search realizes the wishes of the user even before they can understand them, and this is a value for all the current search tools. Besides, these tools give you the possibility to establish a commitment after intelligence models expands its contents.

Maturity Level (1-5)

The overall maturity of the prototype is 3. This prototype is the result of a hard work on innovation through the development of several semantic tools and the aggregated work of the QualiPSo team.

Contact


Julian Pérez – TID
Email: