Mapping contexts to vocabularies to represent intentions Rallou Thomopoulos1 and Marie-Laure Mugnier2 and Michel Leclère3 Abstract. edge representation model based on labelled graphs. The conceptual In the framework of multi-target use of a given ontology, this paper graph model is composed of two parts: the support, which contains proposes a representation of vocabularies based on the identification the terminological knowledge – and constitutes a part of the domain of elementary vocabularies, which can be equivalently defined using ontology –, and the conceptual graphs, which contain the assertional specializations of the “kind of” relation. It defines a way of combin- knowledge. Figure 1 shows a part of the set of concept types, noted ing contexts and vocabularies that allows context-specific querying. TC , which is part of the support. 1 INTRODUCTION A given assertion holds in a given “context”. This single affirmation can be interpreted in various ways, leading to a disparate literature about contexts. We can note two main considerations: (i) a given as- sertion can lead to several interpretations due to different meanings of terms, depending on the context [1, 4, 8]; (ii) the same interpreta- tion can have different truth values in different contexts [9, 6, 12]. In this paper, our concern is to represent that, for the same piece of information, different descriptions will be given, different aspects will be highlighted, depending on the context, which can be seen as the target the piece of information will be used for (for which pub- lic and/or in which purpose). That is to say, different assertions will be used to describe the same piece of information, not because of Figure 1. Part of the “food science” concept type set the ambiguity of terms, nor due to the relativity of truth, but because different aspects will be important to retain, depending on the inten- tion of the message vehiculated in each context. As a consequence, the vocabulary used in each context should be appropriate. Not all A way of representing contexts in this model by structuring knowl- terms of the domain ontology are in accordance with the purposes edge into levels has been descriptively introduced by [11] and fur- of a given context: the presence of unappropriate terms, that do not therly studied e.g. in [3, 10]. The formalization of [2] defines a log- conform to the intended use of information, can reveal a possible di- ically founded knowledge representation formalism based on nested version out of the scope of the context, and thus not be pertinent, graphs, thus providing operations for reasoning with nested graphs. not understandable or not useful. For example, information intended At first level, a conceptual graph gives an overall description of for general public should not be too technical, terms that translate a a fact. Zooming in on certain concept vertices provides more de- judgement (positive, bad, ...) are expected in evaluation contexts, etc. tails, also described by conceptual graphs. A conceptual graph that The aim of this paper is to propose a way of representing vocab- is nested in a concept vertex is thus described in the context defined ularies and associating them with contexts. The examples, although by this concept. Typed nestings [2] allow specifying the relationship simplified, come from a real-world application in food science. The (description, explanation, ...) between the surrounding vertex and one paper is built as follows. Section 2 presents related work on con- of its descriptions. A new type set is thus added to the support, the texts and ontologies. Section 3 defines the proposed representation set of nesting types. In the following, a context is considered to be of vocabularies. Section 4 proposes a mapping between contexts and represented as a nesting type and expresses the target (public and/or vocabularies and shows context-specific querying that ensues. purpose) the nested piece of information is intended for. An example of nested conceptual graphs, built using the concept type set of Figure 1, is given in Figure 2. It represents the following 2 RELATED WORK piece of information: “an article, whose subject is a wheat food prod- 2.1 Context representation uct that is cooked in water, has a result, whose nutritional observa- tion is that the vitamin content of this wheat food product decreases, The context model we use is based on the definition of contexts as whose biochemical explanation is that this wheat food product con- nesting types [2] in the conceptual graph model, which is a knowl- tains hydrosoluble vitamin that is dissolved, and whose nutritional 1 INRA (IATE Joint Research Unit) / Associate Researcher of LIRMM, evaluation is that the nutritional quality of this wheat food product is Montpellier, France, email: rallou@ensam.inra.fr deteriorated”. 2 LIRMM, Montpellier, France, email: mugnier@lirmm.fr The set of conceptual graphs is partially pre-ordered by the spe- 3 LIRMM, Montpellier, France, email: leclere@lirmm.fr cialization relation (noted ≤), which can be computed by the pro- As mentioned in previous works (see part 2.2), in practice ontolo- gies are constructed by successive specializations from top to bottom level. Moreover considering that several direct specializations of a given concept type can have related meanings seems sensible. To conserve these notions, we consider that vocabularies are composed of elementary vocabularies built by successive specializations, in a top-down way, of the concept type set. Definition 2 TC is partitioned into a set of elementary vocabularies Vi built as follows: - V0 is composed of the Universal concept type; - For n > 0, Vn is obtained by defining specializations of concept Figure 2. An example of nested conceptual graphs types of one elementary vocabulary Vk (k < n), or common special- izations of several given elementary vocabularies, through a given specialization criterion4 (noted crt). jection operation (a graph morphism allowing the restriction of the An example is given in Figure 3 for a small part of the set of con- vertex labels). The projection is a ground operation in the concep- cept types. The criterion used for each vocabulary is noted in brack- tual graph model since it allows the search for answers, which can be ets. In this example, each elementary vocabulary is built by special- viewed as specializations of a query (see Section 4.2). izing one preceding elementary vocabulary. 2.2 Ontology structure The question of combining different vocabularies is a major concern of ontology integration. Several studies (e.g. [7, 5]) have proposed distinguishing between different kinds of terminologies according to their level of generality, the top-level being usable for large commu- nities of users, whereas the more specific ones are obtained by spe- cializing the more general levels and used for more specific needs. However, pertinent vocabulary, for a given use, does not always depend on its depth in the ontology. An example is the following. To Figure 3. Example of vocabulary construction express information intended for a general public, we can note that, besides top-level concept types (see Figure 1), several other concept types are pertinent because they correspond to commonly used cate- gories (Spaghetti, Lasagna ...), although they are more specific than Vocabularies can then be built as unions of elementary vocabu- concept types that correpond to technical categories (Extruded pasta, laries, obtained through specialization criteria that make sense for Laminated pasta ...) and hence cannot be used. In this example, this a given informational purpose (see Section 4). The use of the same is due to the fact that Spaghetti or Lasagna are appellations, they do specialization criterion in the definition of different elementary vo- not explicitely express technical criteria. cabularies (for instance in Figure 3, vocabularies V3 and V5 ) can ex- plain why categories that are at different depths in the ontology may be pertinent for the same uses. 3 VOCABULARY REPRESENTATION Due to this consideration, an alternative basis to characterize perti- 3.2 An equivalent definition nent vocabulary for a given use, other than its depth in the ontol- The main idea being that the depth in the ontology is not so important ogy, seems coherent to us. We propose a construction of vocabularies as the specialization criterion, we propose to formalize the notion of based on the specialization criteria used to obtain the concept types criterion as a specialization of the “kind of” relation. that compose them (appellation, technology, ...). We will firstly de- fine “vocabularies”, then propose two equivalent ways of construct- Definition 3 A specialization of the “kind of” relation (noted