Nathalie Hernandez

A semantic web interface using patterns: the SWIP system

Our purpose is to hide the complexity of formulating a query expressed in a graph query language such as SPARQL. We propose a mechanism allowing queries to be expressed in a very simple pivot language, mainly composed of keywords and relations between keywords. Our system associates the keywords with the corresponding elements of the ontology (classes, relations, instances). Then it selects pre-written query patterns, and instanciates them with regard to the keywords of the initial query. Several possible queries are generated, ranked and then shown to the user. These queries are presented by means of natural language sentences. The user then selects the query he/she is interested in and the SPARQL query is built.

A Model to Represent the Facets of Learning Object

This paper presents a model to describe Learning Objects (LO). The main objective of this model is to consider all the aspects of the LO for which a description will ease LO re-use. The LO description we promote complies with the current standards of e-learning and includes the following: metadata, scenarios the objects are used in, and the objects they are composed of. We enrich this multi-facet representation by taking into account the semantics of learning object contents. Another contribution of our work is that this multi-facet representation relies on ontologies, allowing a semantic representation that facilitates communication between machines and users.

A Multi-Agent System for Dynamic Ontologies

In the article, we present Dynamo (an acronym of DYNAMic Ontologies), a tool based on an adaptive multi-agent system to construct and maintain an ontology from a domain-specific set of texts. The originality of our proposal is that the adaptative multi-agent system is used both to represent the ontology itself and to produce the ontology. This enables us to propose a system building and maintaining dynamically an ontology according to interactions with the user (also called the ontologist). We present our system and the mechanisms used to build and maintain the ontology from the texts and for the interactions with the ontologist. We also give results of the evaluation of our system.

An easy way of expressing conceptual graph queries from keywords and query patterns

Our goal is to hide the complexity of formulating a query expressed in a graph query language such as conceptual graphs. We propose a mechanism for querying a collection of documents previously annotated by means of conceptual graphs. Our queries are based on keywords given by the user. Our system associates the keywords with their corresponding concepts. Then it selects a set of patterns which are pre-written typical queries. The patterns are instantiated according to the initial keywords. Finally, the user can modify the pre-written query he/she selects in order to fit his/her initial needs as well as possible. The query graph obtained is projected into the annotation graphs associated with the documents. The documents which answer the query are returned to the user.

Expressing conceptual graph queries from patterns: how to take into account the relations

Our goal is to hide the complexity of formulating a query expressed in a graph query language such as conceptual graphs. We propose a mechanism allowing one to express queries in a very simple pivot language, mainly composed of keywords and relations between keywords. Our system associates the keywords with the corresponding elements of the support (concept types, relation types, individual markers). Then it selects pre-written query patterns, and instanciates them with regard to the keywords of the initial query. Several possible queries are shown to the user. These queries are presented by means of natural language sentences. The user then selects the query he/she is interested in. The query conceptual graph is then built.

Ontologies pour l'aide à l'exploration d'une collection de documents

In this paper we present an ontology-based system to help in the search, analysis and mining of document corpus dealing with a scientific domain. The ontologies used aim to represent a domain both though the vocabulary of this domain and the meta-data that can be needed in accessing information and technological monitoring. The document representation model is therefore built on two different ontologies: an ontology of the domain linked to the task to achieve (here technological watch) and an ontology of the domain dealt with in the corpus. The application field we chose to illustrate our approach is astronomy. The prototype of the system interface is also presented. This interface allows a user to go through the data and to mine the document collection thanks to the representation of knowledge linked to the task and knowledge of the domain.

IoT-O, a Core-Domain IoT Ontology to Represent Connected Devices Networks

Smart objects are now present in our everyday lives, and the Internet of Things is expanding both in number of devices and in volume of produced data. These devices are deployed in dynamic ecosystems, with spatial mobility constraints, intermittent network availability depending on many parameters e.g. battery level or duty cycle, etc. To capture knowledge describing such evolving systems, open, shared and dynamic knowledge representations are required. These representations should also have the ability to adapt over time to the changing state of the world. That is why we propose IoT-O, a core-domain modular IoT ontology proposing a vocabulary to describe connected devices and their relation with their environment. First, existing IoT ontologies are described and compared to requirements an IoT ontology should be compliant with. Then, after a detailed description of its modules, IoT-O is instantiated in a home automation use case to illustrate how it supports the description of evolving systems.

Swip: A Natural Language to SPARQL Interface Implemented with SPARQL

The Swip approach aims at translating into SPARQL queries expressed in natural language exploiting query patterns. In this article, we present the main module of the prototype implementing this approach which entirely relies on SPARQL. All steps of the interpretation process which are carried out in this module are indeed completely performed on RDF triple stores through SPARQL updates. Thus, the implementation gets benefit from SPARQL engine capabilities, which prevent us from worrying about graph manipulation and matching.

Méthodologie de transformation d'un thesaurus en une ontologie de domaine

Information Retrieval techniques make use of terms that are automatically extracted from documents; these terms are used to give information access. In this paper we propose an approach to enrich semantically this extraction by adding knowledge from thesauri. More specifically, the methodology we promote in this paper aims at transforming a thesaurus into a domain ontology which will then be used to semantically index documents (indexes are concepts rather than terms). We also propose techniques that implement this transformation as well as an evaluation in the field of astronomy.

Taking SPARQL 1.1 Extensions into Account in the SWIP System

The SWIP system aims at hiding the complexity of expressing a query in a graph query language such as SPARQL. We propose a mechanism by which a query expressed in natural language is translated into a SPARQL query. Our system analyses the sentence in order to exhibit concepts, instances and relations. Then it generates a query in an internal format called the pivot language. Finally, it selects pre-written query patterns and instantiates them with regard to the keywords of the initial query. These queries are presented by means of explicative natural language sentences among which the user can select the query he/she is actually interested in. We are currently focusing on new kinds of queries which are handled by the new version of our system, which is now based on the 1.1 version of SPARQL.