Top Level Ontology

From D4Science Wiki
Revision as of 13:27, 29 October 2014 by Yannis.marketakis (Talk | contribs)

Jump to: navigation, search

Introduction

In a nutshell

One of the main characteristics of biodiversity data is its cross-disciplinary feature and the extremely broad range of data types, structures, and semantic concepts which encompasses. Moreover, biodiversity data, especially in the marine domain, is widely distributed, with few well-established repositories or standard protocols for their archiving, access, and retrieval. Queries like “Given the scientific name of a species, find its predators with the related taxon-rank classification and with the different codes that the organizations use to refer to them", cannot be formulated (and consequently nor answered) by any individual source. To formulate such queries we need an expressive conceptual model, while for answering them we also have to assemble pieces of information stored in different sources. To fill this gap, we have designed and implemented a top level ontology, called Marine Top Level Ontology (for short MarineTLO).

Motivating Scenarios

The availability of a top level ontology for the marine domain would be useful in various scenarios. Below we will describe them.

For Publishing Linked Data: There is a trend towards publishing Linked Data; consequently a rising issue concerns the structure that is beneficial to use during such publishing. The semantic structure that will be presented can be used by the involved organizations for anticipating future needs for information integration, and thus alleviating the required effort for (post) integration.

For Generating Fact Sheets: FactSheetGenerator is an application provided by IRD aiming at providing factual knowledge about the marine domain by mashing-up relevant knowledge distributed across several data sources. Currently, FactSheetGenerator uses only ECOSCOPE and related knowledge stored in other sources (e.g., about commercial codes or taxonomic information) cannot be exploited. MarineTLO could be exploited for advancing this application, i.e., for providing more complete semantic descriptions.

For Semantic Post-Processing of the Results of Keyword Search Queries: Another big challenge nowadays is how to integrate structured data with unstructured data (documents and text). The availability of harmonized structured knowledge about the marine domain can be exploited for a semantic post-processing of the search results (over dedicated or general purpose search systems). XSearch is a meta-search engine that offers semantic post-processing of search results and is able to analyze the returned results by exploiting also the availability of semantic repositories (e.g. SPARQL endpoints). Xsearch could exploit MarineTLO for providing more complete information about the identified entities.

For Enabling Complex Query Services over Integrated Data: MarineTLO can be used as the schema for setting up integrated repositories that offer more complex query services, which cannot be supported by the individual underlying sources. In general, there are two main approaches for building and querying such repositories: the materialized integration approach (or warehouse approach), and the virtual integration (or mediator) approach (more information about these sources can be found in REF_XXX). The key point is that in both cases we need a schema and MarineTLO can serve this requirement.

MarineTLO as a product

...

TLO-Development activity

General Description

This activity concerns with the development of a top level ontology (called MarineTLO) that will integrate the concepts currently existing in marine-domain knowledge bases (in particular FLOD and ECOSCOPE knowledge bases). The MarineTLO-development activity is dived into six sub-activities (or Tasks) and related to each other as shown in the diagram in Fig 1.


Pic1.png

Methodology

The methodology is based on an Iterative and Incremental development approach. As such, one iteration will involve all the tasks of the Fig 1. that are described @ http://wiki.i-marine.eu/index.php/Top_Level_Ontology. All the iterations will be accurately described and evolution of the MarineTLO ontology will be released in each iteration, based on the acquirement of the specific marine domain knowledge.

Activities scheduled with deadlines

Each iteration is planned to be monitored by opening related tickets.

Related Cluster

http://wiki.i-marine.eu/index.php/Semantic_cluster_achievements

Related Wiki Pages

http://wiki.i-marine.eu/index.php/XSearch


Motivation - Goal - Requirements

At very abstract level we use the picture below as preview that focuses on viewpoints for motivations, goal and requirements of having such a ontology on top of marine-domain knowledge bases.

Motivation. Semantic technologies, applications and services for biodiversity mostly rely on the rise of an interconnected and shared tree-of-life like dataset scaling on the web. The various communities (including also marine one) are contributing to this joint effort aim to share domain data and their meaning, to provide a solid basis for biodiversity systems interoperability. One of the challenges in iMarine is how users could experice a coherent source of fact about marine resources, rather than a bag of contributed contents.

Goal. The goal in modelling and formalising MarineTLO ontology is for integrating and semantically extending the underlying models of existing marine data sources. Specifically, the MarineTLO is used on the top of a number of real and heterogeneous marine data sources, including FLOD and ECOSCOPE, as knowledge mediator to represent, manipulate and reason upon and across them.

Requirements. Focusing on Ecosystem Approach to Fisheries and Marine Resources, the MarineTLO ontology should be generic enough

  • to provide consistent abstraction or specification of concepts included in the data models or ontologies of marine data sources
  • to provide the necessary properties to make this distributed knowledge bases (FLOD, ECOSCOPE and others) a coherent source of facts, relating observation data to the respective space-temporal context and categorical domain knowledge


Slide15.jpg

Related Meetings (and their slides)

TCOM in Italy, Rome, 03.11.2012

FLOD Ontological Analysis, 7.12.2012, meeting online

TCOM in Belgium, Ostende, 29.01.2013

Semantic Cluster Meeting, online, 13.02.2013

TCOM in Italy, Pisa, 21.03.2013

TCOM in Italy, Rome, 26.03.2013

TCOM in Greece, Skiathos, 18.06.2013 (LATEST)

Analysis, Design and Implementation

FLOD Ontological Analysis

This activity has the primary goal to provide a common understanding of the underlying model of the FLOD source. It has been considered necessary for the development of the MarineTLO. The associated ticket is https://issue.imarine.research-infrastructures.eu/ticket/888

ECOSCOPE Ontological Analysis

This activity has the primary goal to provide a common understanding of the underlying model of the ECOSCOPE source. It has been considered necessary for the development of the MarineTLO. The associated ticket is https://issue.imarine.research-infrastructures.eu/ticket/889

MarineTLO Design

The activities is related to the design of MarineTLO ontology. The associated ticket is https://issue.imarine.research-infrastructures.eu/ticket/890

MarineTLO Implementation

The activities is related to the implementation of MarineTLO ontology using OWL 2 language. The associated ticket is https://issue.imarine.research-infrastructures.eu/ticket/891

MarineTLO Usage

This activity concerns with the identification of use cases motivating the need for having harmonized integrated information. It has been associated to the ticket https://issue.imarine.research-infrastructures.eu/ticket/900. Currently, we evaluate the MarineTLO ontology for the:

  • Fact Sheet Generator. In this case, MarineTLO is used as top model for FLOD, ECOSCOPE and WoRMS to support the development of FactSheetsGenerator applications. An concrete example is the one provided by IRD (SpeciesFactSheetsGenerator) aiming at providing factual knowledge about the marine domain by mashing-up relevant knowledge distributed across several data sources.
  • For Semantic Post-Processing of the results of keyword search queries. In this case, MarineTLO is used as knowledge model for semantic search in X-search meta-engine. In more detail, suppose that a user is looking for publications about tuna. Specifically he wants to find experiments that were applied to several species of tuna. So, he submits the query tuna and gets a sorted list of results and various categories of entities like Regional Fisheries Body, Species, FAO Country, etc. User realizes that the category Species may contain interesting entities. He notices that there is an entity with the label skipjack tuna which is a medium-sized fish in the tuna family found in tropical and warm-temperate waters. User wants to learn more information about that species. Specifically, he would like to see other species for which the skipjack tuna is predator or is prey. By clicking the icon next to the entity's name, user is able to instantly (at real-time) retrieve such information. In particular, in the back end, a SPARQL query is sent to the MarineTLO's endpoint asking for that information. Note that the 'Species' have been derived from FLOD, while the properties 'is predator of' and 'is prey of' have been derived from ECOSCOPE's knowledge base. That would be impossible without the exploitation of the MarineTLO.

MarineTLO Evaluation

This activity is related to the evaluation of the MarineTLO ontology. The associated ticket is https://issue.imarine.research-infrastructures.eu/ticket/892

A required activity for the MarineTLO evaluation is to populate it with concrete instances.

MarineTLO-related Products

The Notion of MarineTLO Version

Each MarineTLO version consists of

  • A release number
  • OWL files
  • A document that contains the scope notes of each class
  • A set of competency queries
  • A short description describing the changes
  • It could also contain a set of mappings between data source and MarineTLO, each of them described as an OWL file

Evolution Process

Since last meeting in Rome (26-03-2013), we planned to release a new version every two months (for correcting errors, based on requirements, priorities, usage needs, etc).

Version 1.0.0 released on 26-03-2013

Version 2.0.0 planned to be released by end of July 2013

Version 3.0.0 planned to be released by middle of October 2013


Version 4.0.0

  • Release
    • Version: 4.0.0
    • Date: July 2014

Previous TLO Versions

MarineTLO-based Warehouses

Warehouse 1 (June 2013)

Warehouse 2 (By the end of July 2013)

The warehouse is available at http://62.217.127.213:8890/sparql and the graph name is <http://www.ics.forth.gr/isl/TLObasedDataWarehouseV2>. We have also installed some plugins for browsing the warehouse, specifically one can use http://62.217.127.213:8890/fct/ and http://62.217.127.213:8890/fct/demo_queries.vsp that has some demo queries (use “Run with iSPARQL”, because the “Run in SPARQL endpoint” plugin prunes whatever follows the ‘#’ character). The total number of results this plugin returns is limited to 50.

A summary of the warehouse contents follows:

  • TLO version 2
  • FLOD
  • ECOSCOPE
  • part of WoRMS (information about the taxonomies of approximately ~1100 species obtained through Species Discovery Service and the wrapping software developed)
  • (marine) part of DBpedia (containing various information about marine species)

In numbers, this warehouse contains approximately 1.6 million triples about 19,000 distinct marine species.

A paper describing the process of creating the MarineTLO-based warehouse can be found here.

EVALUATION and USAGE

The new warehouse has been evaluated using a new set of competency queries.

X-Search now uses this warehouse (instead of FLOD) and can identify 25,000 marine species (this number includes species genera and family names). Furthermore, each species in the TLOMarine-based warehouse has in average 30 properties, while in FLOD each species has in average only 6 properties.

We are also in continuous collaboration with IRD who is testing (and provides requirements) for the needs of FactSheetGenerator.

FUTURE

Continuous inspection of the warehouse contents, exploitation issues (new queries, etc.), documentation. A next version of the warehouse will be constructed certainly when we reach Marine TLO v3 (scheduled for Oct 2013) based on the requirements that we continuously receive from IRD.

Warehouse 3 (ONGOING, by the end of October 2013)


Warehouse 3+ (ONGOING, by the middle of January 2014)

  • Virtuoso Repository
  • NameGraph
  • Virtuoso Repository Browsing

MarineTLO Related Tickets

First Iteration

Second Iteration

Third Iteration


Fourth Iteration

Related Papers

  • Y. Tzitzikas, C. Allocca, C. Bekiari, Y. Marketakis, P. Fafalios, M. Doerr, N. Minadakis, T. Patkos, and L. Candela , “Integrating Heterogeneous and Distributed Information about Marine Species through a Top Level Ontology”, 7th Metadata and Semantics Research Conference, MTSR'13, Thessaloniki, Greece, November 2013.
  • Y. Tzitzikas, C. Allocca, C. Bekiari, Y. Marketakis, P. Fafalios, and N. Minadakis, "Ontology-based Integration of Heterogeneous and Distributed Information of the Marine Domain", ERCIM News 2014 (96), Special theme: Linked Open Data, January 2014.
  • Y. Tzitzikas, N. Minadakis, Y. Marketakis, P. Fafalios, C. Allocca, and M. Mountantonakis, "Quantifying the Connectivity of a Semantic Warehouse", 4th International Workshop on Linked Web Data Management, LWDM'14, Athens, Greece, March 2014.
  • M. Mountantonakis, C. Allocca, P. Fafalios, N. Minadakis, Y. Marketakis, C. Lantzaki, and Y. Tzitzikas, "Extending VoID for Expressing the Connectivity Metrics of a Semantic Warehouse", 1st International Workshop on Dataset Profiling & Federated Search for Linked Data (PROFILES'14), in conjunction with the 11th Extended Semantic Web Conference (ESWC'14), Anissaras Hersonissou, Crete, Greece, May 2014.
  • Y. Tzitzikas, N. Minadakis, Y. Marketakis, P. Fafalios, C. Allocca, M. Mountantonakis, and I. Zidianaki, "MatWare: Constructing and Exploiting Domain Specific Warehouses by Aggregating Semantic Data", 11th Extended Semantic Web Conference (ESWC'14), Anissaras Hersonissou, Crete, Greece, May 2014.
Retrieved from "https://wiki.d4science.org/index.php?title=Top_Level_Ontology&oldid=20850"