A Comparison of Approaches to Model Uncertainty in Time Intervals

Abstract

Information systems model parts of reality by representing properties of real-world objects or concepts. As real objects or concepts often have temporal aspects, temporal notions such as time intervals are often represented. However, these may contain imperfections like uncertainties, complicating their representations. A very important purpose of information systems is to be able to query them to retrieve information, but representations of temporal notions containing uncertainties severely complicate querying. Thus, several soft computing techniques have been proposed to represent time intervals subject to uncertainties in a semantically sound way and to reason with them in a semantically sound and useful way. In the presented work, two frameworks designed for this are compared. It is found that, despite slight differences in the way these frameworks represent intervals, they provide the same results when reasoning about time intervals subject to uncertainty.

Combining quantifications for flexible query result ranking

“…Often, a set of data exists, each of which could be the intended datum. In those cases, it is said that the datum is subject to uncertainty [1], [2], [5], [8], [12], [21], [24]- [27]. Usually, for a datum subject to uncertainty, an attempt is made to model the available knowledge about the intended datum, by assigning each of the data which could be the intended datum a degree of confidence an agent has that the corresponding datum is the intended datum.…”

“…In this paper, a quantification is exactly this: a numerical value expressing a valuation. Thus, a similar data collection is again considered a query answer, but, compliance to the user's preferences is now a matter of degrees (quantifications) [2]- [7], [10]- [13], [15]- [17], [19]. For example, the result set of the last example query applied to the example relation visualized in table I might take the form of the set of tuples visualized in table II.…”

Combining Uncertainty and Vagueness in Time Intervals

“…Through such data (a.o. ), DBS model these real-life objects or concepts [2], [14]. For example, patient DBS may model patients' medical state by containing data representing these patients' heart rate, blood pressure, ...…”

“…For example, patient DBS may need to model patients' medical history by containing data representing the time intervals during which the patients suffered from certain diseases or data representing the instants when the patients took medication. Because time intervals in DBS are studied more exhaustively in literature [1], [2], [5], [8], [12], [14], [15], [16], [17] than instants in DBS [7], [8] and because an instant can be seen (and modeled) as a time interval, the work presented in this paper only considers time intervals in DBS.…”

“…Such data represent properties which can be the result of measurements, estimations, derivations, ... However, human-made data, information and knowledge can be prone to imperfections, due to the inherent imperfect nature of humans and human reasoning or imperfections in measuring equipment [1], [2], [5], [6], [7], [8], [9], [11], [12], [13], [14], [15]. For example, a medical professional may need to estimate the exact moment in time when a patient was given a certain medicine, as nobody remembers exactly, or the period of time during which a patient suffered from a certain disease, fully knowning that the emergence and the disappearance of the disease are gradual.…”