GoNTogle is a semantic annotation and search tool, built on top of Lucene and Protege. It supports manual and automatic annotation of several types of documents (doc, pdf, rtf, txt, odt) using ontology concepts. It also provides searching facilities beyond the traditional keyword-based search, including (advanced) ontology-based search, and combined search.
What does GoNTogle offer?
- It allows users to open and view widely used document formats such as .doc and .pdf maintaining their original format.
- It provides an easy and intuitive way of annotating documents (or document parts) using ontologies.
- It provides an automatic annotation mechanism based on user history, thus, offering personalization of annotations.
- It is based on a server-based architecture, where document annotations are reachable from all users.
- It combines keyword and ontology search, providing advanced search facilities for both types of search.
Click next figure to see GoNTogle in use (for more details check some screenshots and a use case here).
GoNTogle is described in a demo and a full research paper:
- Integrating Keywords and Semantics on Document Annotation and Search, (ODBASE'10).
- GoNTogle: a Tool for Semantic Annotation and Search, (ESWC'10) (Demo).
There are, also, 2 diploma theses describing in more detail the development of GoNTogle:
- Automatic Categorization and Semantic Query Interpretation for Ontology Search (in greek), A. Kourti, 2008.
- GoNToggle: Smart searching using ontologies (in greek), G. Giannopoulos, 2006.
The following screenshots and video demostrate several examples of GoNTogle use.
Manual Annotation. GoNTogle provides two annotation options:
1. Annotate the whole document: the user opens a document, loads an ontology, selects the ontology class from the Protégé Frame and chooses "Create New Annotation On Whole Document", as shown in the following figure.
2. Annotate parts of document: the user opens a document, highlights a part of the document, selects the ontology class from the Protégé Frame and chooses "Create New Annotation On Selected Text", as shown in the following figure.
Automatic Annotation. By activating the automatic annotation, the system suggests ontology classes to be used for annotation. For example, in the following figure, we can see that the system suggests the class H.2_DATABASE_MANAGEMENT as the most appropriate class for the current document.
Querying facilities. Next we present some example use cases, where several types of queries are posed in GoNTogle:
Keyword based queries. Find documents relevant to XML. Next figure illustrates the operation.
Ontology based queries. Find documents annotated with the class H.2_DATABASE_MANAGEMENT. Next figure illustrates the operation.
Keyword/Ontology based queries. Find documents relevant to XML annotated with the class H.2_DATABASE_MANAGEMENT. Next figure illustrates the operation.
Get next generation search. ssume that the user searches for documents relevant to XML annotated with the class H.2_DATABASE_MANAGEMENT. Next figure illustrates the operation. The previous figure illustrates this operation. User has then the option to propagate the search to lower levels in the ontology (i.e., searching in all subclasses of H.2_DATABASE_MANAGEMENT) by using the Get Next Generation option. Next figure illustrates documents retrieved using the Get Next Generation option.
Binaries. You can download GoNTogle here.
- Windows Installation:
- Unzip the gontogle.rar into a folder of your choice.
- Download and instal openoffice 3.
Add to the system variable 'Path' the following 2 paths:
- "path to openoffice installation directory"\Basis\program (e.g.: C:\Program Files\OpenOffice.org 3\Basis\program)
- "path to openoffice installation directory"\URE\bin (e.g.: C:\Program Files\OpenOffice.org 3\URE\bin)
Create a new system variable named 'UNO_PATH' and add the following path:
- "path to openoffice installation directory"\program (e.g.: C:\Program Files\OpenOffice.org 3\program)
- Restart your computer.
After the above steps are completed, you can run gontogle either by double-clicking on the dist\goNTogle.jar or by running
- java -jar "path_to_gontogle_installation_directory\goNTogle.jar"
Setup. The following instructions describe the gontogle setup process that is necessary to take place after the installation. Go to menu Tools-> Options:
Open Office folder setup.
- Set the Open Office installation folder in the Open Office field.
To search/annotate a document collection you should first set up the indexes. Click on Create Index.
- Set the folder containing the documents to be indexed.
- Set the folder to store the index.
- To add documents to an already existing index, select Select an already existing Index to add new elements and set up the folder where that documents exist.
- To select an already existing index, click on Selected index and make your selection.
- To search/annotate a document collection you should first set up the indexes. Click on Create Index.
Ontology Server setup.
- Set the IP address of the protege ontology server. The default setting is the IP of our server.
- Open Office folder setup.