3 |
State of the Art in Temporal and Cultural Ontologies
The aim of this chapter is to twofold: First is to mention what kind of ontological models exist for the domain of time and for the domain of culture within the context of the Web. Second is to discuss whether socio-cultural time ontologies exist. If they do not exist, it shall be discussed how the current ontologies address socio-cultural conception of time. Here, we assume that a number of time and culture ontologies exist in the field of multiagent systems, yet we will concentrate on the Web ontologies.
During our research we have come across to numerous independent domain ontologies of time for the Web, which are in different ontology languages and which have different conceptual models e.g. [13],[39]. Simply sending the query “time ontology” to the search engine Google returns a vast quantity of time ontologies. Additionally, as time is a general notion of reality like space, process, event and so forth, it has been modelled as a part of various commonsense ontologies such as the Cyc ontology and others.
Concerning culture, some ontologies exist that cover this domain in its broadest sense. Ontologies about cultural heritage systems [48],[21],[17],[78], ontologies about cultures in different enterprises and organisations [6], and ontologies about culture-specific language terms [7] are examples. Nevertheless, we have not come across to any ontology of socio-cultural time that formally describes the socio-cultural conception of time.
One of the most well-known ontologies about time is the Time-DAML ontology[1], which is named after the ontology language it is written in. As such, it is an outcome of the collaborative efforts of a number of researchers, whose motivation has been to develop an ontology to facilitate the adoption of a single representative ontology for the domain of time. The purpose of the Time-DAML ontology is to deliver formal temporal information about the contents of Web resources.
Very briefly, Time-DAML partitions the domain of time into two categories of time points and time intervals and defines relationships that hold between the time intervals based on Allen’s temporal interval calculus [8]. Defined time units include seconds, minutes, hours, days, months and years. Events are not described internally; instead some relations are defined to link the Time-DAML ontology to external ontologies that describe events [39]. The reason why we are mentioning Time-DAML ontology here is that we believe it falls short in terms of addressing the conception of socio-cultural time. Let us give an example from the Time-DAML reference to present the description of (socio-cultural) time concepts like weekendday and weekday in this ontology: An extract of Time-DAML ontology
Table 1 Definition of weekday and weekendday in Time-DAML ontology
According to this description, a weekday is defined in an enumerative manner to be one of the days of Monday, Tuesday, Wednesday, Thursday or Friday. In the same manner, the definition of weekendday concerns one of the two days of Saturday or Sunday. Although this definition may hold true for many cultures, it does not apply for a number of other cultures. For example, in Turkey, Friday may also be considered as a weekendday. Although Turkey officially uses the Gregorian calendar and the official weekend days are Saturday and Sunday, many observing Turkish businessmen do not work on Fridays and concieve of Friday as a weekendday. Yet another example is Israel, where Sunday is not a weekendday but it is the first weekday. Clearly, the current time model of Time-DAML does not refer to the culture-dependent conception of time.
Ontologies related to the domain of culture are not as high in number as the ontologies of time. Also, most of them concentrate on the domain of cultural heritage, in other words they model data related to museums. The ontology of CIDOC Conceptual Reference Model for cultural heritage systems is the most well-known ontology in this field [48]. It is a domain ontology that has an object-oriented model and that has the purpose to solve the problem of semantic inter-operability between museum data. In the ontology of CIDOC Conceptual Reference Model for cultural heritage systems, the concept of time is explicitly addressed. However, it is not related to the conception of socio-cultural time. Temporal information in CIDOC model is referred to by means of the categories TemporalEntity and TimeSpan. For example, “Impressionism” can be an example of a TemporalEntity and the time extent of “from 1867 to 1886” can be an example of TimeSpan. Then, by means of these and other primitives, it becomes possible to state the physical condition of a museum artefact, be it a painting, a piece of multi-media, a book or any other object, and its associated time period. A detailed illustration of the model can be found at [58]. However, this model is silent about the notion of socio-cultural time as it handles a completely different relation between culture and time, namely the associated time periods of the artefacts of cultural heritage.
Another ontology about culture is the one conceptualized by Abou-Zaeid in [6] and his ontology formally describes different cultural values within the context of international business enterprises. Drawing upon the assumption that today there are many international enterprises that work colloboratively over the borders, the purpose of his ontology is to build a shared vocabulary that explicitly specifies the similarities and differences between their value systems.
In short, Abou Zaeid’s ontology is intended to facilitate the inter-cultural communication with regard to business enterprises. His cultural ontology focuses on inter-organizational processes, therefore the concept time is present in the ontology as an inseparable part of any process. However, as we have stated, the notion of socio-cultural time remains unaddressed. Otherwise, the cultural ontology defines two concepts that are organizational-specific values and relevant national-specific values. National-specific value is defined as “a common value shared by people within a given country and can be used to distinguish one country’s culture from another” [6]. Although such a definition may be argued to contain the conception of time as a common value for a nation or a culture in a very general manner, there is no explicit reference to the perception of culture- or nation-specific time.
To conclude, we can state: as a result of our research, we have observed that separate domain ontologies model the notion of time and culture independently from each other. Additionally, culture seems to be a very loosely defined subject, therefore different ontologies model different interpretations of the concept culture. In the end, we have failed to find an ontology that handles the conception of socio-cultural time. Furthermore, we believe current time ontologies and culture ontologies fall short in addressing the notion of socio-cultural time.
[1] http://www.cs.rochester.edu/~ferguson/daml/
|