=Paper=
{{Paper
|id=None
|storemode=property
|title=The Time Dimension in Information Logistics
|pdfUrl=https://ceur-ws.org/Vol-1028/paper-04.pdf
|volume=Vol-1028
|dblpUrl=https://dblp.org/rec/conf/bir/Gaidukovs13
}}
==The Time Dimension in Information Logistics==
https://ceur-ws.org/Vol-1028/paper-04.pdf
The Time Dimension in Information Logistics
Andrejs Gaidukovs and Marite Kirikova
Institute of Applied Computer Systems, Riga Technical University, Latvia
{andrejs.gaidukovs; marite.kirikova}@rtu.lv
Abstract. The purpose of information logistics is to ensure that the right
information, which is necessary in accomplishment of business tasks, is
available in the right location, in the right time, and in the right quality.
Different combinations of models have been suggested for supporting this
purpose. However, none of them includes the model that corresponds to the
time dimension of information logistics. Taking into consideration that the right
time is an essential goal that corresponds to the purpose of information
logistics, this paper takes a closer look at the time dimension and suggests
extending the set of related models of information logistics with the time
dimension model.
Keywords: information logistics, time dimension, information demand model
1 Introduction
Information Logistics is a branch of research that addresses concepts, methods,
technologies, and solutions that help to ensure situation sensitive availability of high
quality information for individuals or groups with respect to their information needs,
time, location, and user-friendly form of representation [1]. As revealed in [2] the
notion of Information Logistics was coined in 1978. According to data available in
scientific repository SCOPUS, the number of papers on information logistics has
gradually increased from one paper in 1982 [3] to more than 40 papers per year
during 2009-2012.
Information logistics usually is organized around the following four main
dimensions [4]:
Personalization: each person or group has particular information needs that
depend on their knowledge and experience
Time: information has to be available or delivered in a particular time when it
is actually needed
Communication: information has to be available (or represented) in the form
that is convenient for its users
Context: information has to be deliverable regarding the location and situation,
in which it is used in a particular moment of time
One more dimension – the quality dimension of information logistics is suggested
in [1].
Most of the research in the information logistics has been devoted to the context
dimension, e.g., [5], [6], [7], and [8]. Usually the time dimension is considered from
the point of view of sequence of time moments or intervals in which particular items
of information are needed. The intervals can be relative, e.g., “four weeks before a
particular time moment”, “six weeks before a particular time moment” and so forth.
The moments and start and end points of the intervals are represented on a scalar time
axis. Thus, while such issues as roles, tasks, information items, and quality parameters
are addressed in specific interrelated models, the time dimension has not been
considered as a separate information logistics model so far.
The goal of this paper is to analyze the time dimension in depth and propose the
model of the time dimension that would enhance possibility to represent and analyze
information needs and information availability and delivery patterns more accurately.
The paper is organized as follows: In Section 2 features of the time dimension are
discussed. In Section 3 conceptual modeling of time dimension is considered and the
model of the time dimension is presented. In Section 4 the time dimension is put in
the context of information demand modeling. In Section 5 the practical applicability
of the time dimension model is discussed. Brief conclusions are stated in Section 5.
2 Features of Time Dimension
In this section the discussion of the time dimension is pragmatically oriented; it is
mainly based on research reflected in [9] and [10] and does not concern philosophical
considerations of time dimension.
The features of time can be structured as reflected in Fig 1.
Fig. 1 Features of time
Time can be considered as discrete items or as a continuous phenomenon. In
information technology context it is mainly considered as discrete items. It can be
positioned as absolute time referring to a chosen “clock” or as a relative time (e.g.,
“one year ago”). An essential feature of time is “periodicity” or frequency of
intervals that can be reflected using linear or non-linear scales. It is essential that users
are free to define time periods by themselves. In related work one can find the
following types or aspects of the time dimension:
Calendar granularity (N years, Year, Half-year, Quarter, Month, Week, Day,
Hour, Minute, Second)
Monthly (January, February, March, April, May, June, July, August,
September, October, November, December)
Daily (Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday)
Part of the day (Day time, Night time)
Seasonal (Spring, Summer, Autumn, Winter)
Arbitrary time intervals (periods)
Arbitrary time moments
Time zone
Process time (Start time, Duration, End time, Deadline)
Relative pointers (Before, After, Now, Every)
Time types (systems time, real time, see e.g. ISO SQL:2011 standard [11])
The model of the time dimension, which is proposed in the next section, includes
most of above-listed time dimension types.
3 The Model of the Time Dimension
While there is rarely specific time model available in the related work, time has been
included in different conceptual models. Ten of such models are surveyed in [9]:
Infologistic data model by Langefors, presented in 1973 [12]. In this model the
time (moment or period) is an attribute of the elementary fact, belonging to its
built-in context. E.g., O(p) is a set of objects with property p, Ot(p) is a set of
objects which have property p during time period t.
Conceptual Information Model [13] distinguishes between extrinsic time
automatically set by the software system and intrinsic time (a part of the fact
that refers to real time when the fact is true). Here the time is considered as a
component of the relationship. In its later modifications the distinction
between time moments and time intervals is taken into consideration.
The time extended Entity Relationship Model – TERM [14] focuses on
historical structure of data, where each attribute, role, or relationship has
historical property. This is achieved by specific basic, derived, or inductive
historical operations.
Logic based model proposed by Lundberg [15], which argues that, despite of
continues nature of time, in information systems it is to be represented as a
discrete phenomenon.
INFOLOG [16] model maintains time dimension as a set of temporal
operators. This model ignores future and focuses only on past and present
issues of facts.
DMILT [17] model is based on specific temporal logic that addresses process
network, where processes exchange messages via the temporal database of the
information system.
The Entity, Relationship, Attribute, Event (ERAE) [18] model distinguishes
between past, present and future of individual data entities or groups of
entities. Time is represented as a specific data type.
The Conceptual Modeling Language (CML) [19]. This language has object-
oriented structure. It considers discrete time reflected as time intervals using
such predicates as meets, equals, during/over, before/after,
startsbefore/startsafter, endsbefore/endsafter, overlaps, costarts, and coends.
This language includes also two constants, namely, Alltime and Now.
TEMPORA [20] consists of two types of models such as Entity Relationship
Time (ERT) model and Conceptual Rule Language (CRL). Time is reflected
as time stamps of entities and relationships in ERT. CRL includes different
predicates that allow reflecting various time based situations. Time moments
and time intervals are considered (for events – only time moments). Modeling
time and real time are considered, as well as historical relationship is
presented.
While above-listed methods have been developed quite a time ago and time is an
essential issue in contemporary business and information systems environment, there
are not many research works available that would consider conceptual modeling
aspects of the time dimension. We can distinguish between the following main areas
where time issues are currently discussed:
Neuropsychology
Simulation
Data Warehouses
Business Intelligence
Ontology Engineering
In neuropsychology, simulation, data warehouses, and business intelligence only
some aspects of the time dimension are considered [21], [22], [23], [24]. The broadest
scope of the aspects is analyzed in the research on the time ontology [25], [26].
The fact that recently issued SQL:2012 standard [11] considers only some aspects
(calendar granularity, real time, and systems time) of the time dimension, shows that
it is necessary to further research the time dimension to better understand its features
and incorporate it into models of information logistics.
In this paper we present the first version of the time dimension model that has been
developed, mainly, by amalgamating various aspects of the time dimension in a single
model. This was done with the purpose to obtain a generic view on different aspects
of time relevant in information logistics and to have a possibility to control
relationships between these aspects in the time dimension model. The simplified
version of the proposed model is presented in Fig. 2. The model presented in Fig. 2
has been obtained independently of the time ontology presented in [26]. While the
proposed model has many similarities with the time ontology, such as inclusion of
parts of the year, time zones, etc; still there are some essential differences how
periodicity and time moments are handled. The model in Fig. 2 distinguishes between
periods and intervals as separate entities since both the length of transactions
(intervals) and the periods of time showing repetitive execution of transactions are
important in information logistics. More detailed comparison of the time ontology and
proposed time dimension model is beyond of the scope of this paper.
Fig. 2 Time dimension model (simplified)
Fig. 2 reflects different aspects of time that can be relevant in various tasks of
information logistics (See Section 2). The positioning of the time dimension model in
the context of other models of information logistics is discussed in the next section.
4 Time Dimension and Information Demand Model
For illustrating the role of the time dimension in information logistics we use
Information Demand Pattern discussed in [2]. From the modeling perspective this
pattern prescribes the model that consists of the following sub-models:
Information Model that represents the items of information used in
performance of tasks
Effects Model that reflects different issues of quality of information logistics,
such as economical effect, motivation, experience etc.
Organizational Chart showing relationships between the roles in organization
Task Model showing the architecture of the tasks.
The elements of above-listed models should be iner-related in order to show, which
information is needed for which role when performing a particular task, and what
effects the availability/unavailability of this information may cause. Fig. 3 illustrates
how these models can be related to the time dimension model.
Fig. 3 Information Demand Model related to the time dimension model
At high level of abstraction Fig. 3 shows how elements of Information Demand
Pattern can be related to the time dimension model. However, any element of the time
dimension model reflected in Fig. 2 can be indirectly related to Information Demand
Pattern. This is illustrated by the time dimension model bellow the dotted line in Fig.
3. The usage of the time dimension affects the Task Model. Each Task Model’s
element has to have attributes that help to characterize their duration (performance
time or interval) and periodicity (period or interval of repetition of the task).
We can distinguish between the following types of tasks:
Meetings: project meetings, conferences, seminars
Ordinary Tasks: performing a particular transformation
On one hand, the type of the task does not change the way how the time dimension
is incorporated in Information Demand Pattern. On another hand, the experimentally
obtained models, which represented particular Information Demand Pattern’s
(extended by the time dimension model) instances, revealed that the extended
Information Demand Pattern models have slightly different outlooks of above-
mentioned types of tasks. Nevertheless, in both cases there is a time interval t that is
related to start time point ts and end time point te of the task
t = te - ts (1)
Moment of Delivery Md is defined as a moment in which a particular information
unit is sent to the role. In most cases this moment should not occur later than the
beginning of the task.
Both types of tasks can occur periodically. In this case it is necessary to define
period P as the delta between two sequential starting points of tasks ts.n and ts.n+1,
P= ts.n+1 - ts.n (2)
The periods can be constant and arbitrary. In case they are constant ones (e.g.
handing in monthly reports), the Md can be calculated as follows:
Md=P-t (3)
Often in information logistics the moment, when information is received, can be
considered being practically equal to the moment, when information is delivered.
However, there are cases, when the time of delivery has to be taken into account.
5 Experimental Application of Time Dimension Model
The time dimension model proposed in Section 2 related to Information Demand
Pattern (Section 3) was applied for information demand modeling in the project
proposal preparation in a public organization. Effects Model was not used in the
experiment. The models illustrating the obtained information demand pattern are
shown in Fig. 4.
Fig. 4 Part of Information Demand Pattern models for project proposal preparation extended by
the time dimension model
There were six organizational roles, nineteen tasks, and twenty five information
units defined. The maximal duration of the project proposal preparation was 25 days.
On the basis of the models, the information system’s infrastructure and the data model
were defined for supporting the tasks represented in the Task Model. The time
dimension model was helpful for identification of time issues and attributes relevant
for information logistics of project proposal preparation. The model helped to
represent and analyze information needs and show information availability and
delivery patterns more accurately.
6 Conclusions
The paper focuses on the time dimension in information logistics. It contributes the
first version of the conceptual time dimension model for information logistics as well
as relates well-known Information Demand Pattern to the time dimension.
First experiments with inclusion of the time dimension model in the set of inter-
related information logistics models show that the time dimension model helps to
represent and analyze information needs and show information availability and
delivery patterns accurately. The model is useful for visualizing the tasks of roles and
designing information architecture that supports the performance of the tasks.
Further research should concern tuning of the time dimension model, lager scale
experiments with Information Demand Pattern extended by the time dimension
model, and designing transformation algorithms that produce role oriented and
information technology support oriented views on the detailed sub-models of the
extended Information Demand Pattern.
References
1. Sandkuhl K.: Information Logistics in Networked Organizations: Selected Concepts and
Applications. Enterprise Information Systems: 9th International Conference, ICEIS 2007,
Funchal, Madeira, June 12-16, 2007, Revised Selected Papers. Springer, pp. 43-54 (2009).
2. Sandkuhl K., Information Logistics, The University of Rostock (2011).
3. Reichertz P.L.: Future developments of data processing in health care. Methods of
Information in Medicine 21 (2), pp. 55-58 (1982).
4. Haseloff S.: Context awareness in Information Logistics, PhD thesis, Technischen
Universität Berlin, Berlin, Germany, 247p. (2005).
5. Sandkuhl K.: Information Demand Patterns, Xpert Publishing Services, Rome (2011).
6. Sandkuhl K., Smirnov A., Shilov N.: Information Logistics in Engineering Change
Management: Integrating Demand Patterns and Recommendation Systems. Workshops on
Business Informatics Research: BIR 2011 International Workshops and Doctoral
Consortium, Riga, Latvia, October 6, 2011 Revised Selected Papers, Springer Berlin
Heidelberg, pp. 14-25 (2012).
7. Michelberger B., Mutschler B., Reichert M.: A Context Framework for Process-Oriented
Information Logistics. Business Information Systems. 15th International Conference, BIS
2012, Vilnius, Lithuania, May 21-23, 2012. Proceedings., Springer Berlin Heidelberg, pp.
260-271 (2012).
8. Sandkuhl K.: Material Specification Responsible, Jonkoping University (2009).
9. Theodoulidis C.I., Loucopoulos P.: The time dimension in conceptual modelling”,
Information Systems, vol. 16, pp. 273-300 (1991).
10. Chendroyaperumal C., Time Dimensions for Managerial Excellence, SSRN:1544363
(2010).
11. Kulkarni K., Michels J.-E.: Temporal features in SQL:2011. SIGMOD Rec.,vol. 41, pp.
34-43, (2012).
12. Langefors B.: Theoretical analysis of information systems, Studentlitteratur., Lund (1973).
13. Bubenko J.: Information Modelling in the context of System Development. IFIP
Information Processing '80, Amsterdam, pp. 395-411 (1980).
14. Klopprogge M.R.: TERM: An Approach to Include Time Dimension in the Entity-
Relationship Model. Proceedings of the Second International Conference on the Entity-
Relationship Approach to Information Modeling and Analysis, North-Holland Publishing
Co., Amsterdam, pp. 473-508 (1981).
15. Lundberg B.: An axiomatization of events. BIT Numerical Mathematics, vol 22. (1982).
16. Sernadas A.: Temporal aspects of logical procedure definition. Information Systems, vol.
5, pp. 167-187 (1980).
17. Carmo J, Sernadas A.: A Temporal Logic framework for a fayered approach to systems
specification and verification. Proceedings of the IFIP 8/WG8.1 Working Conference on
Temporal Aspects in Information Systems, pp. 31-46 (1987).
18. Dubois E., Hagelstein J., Lahou E., Ponsaert F., Rifaut A.: The ERAE Model: A Case
Study. Proc. of the IFIP WG 8.1 Working Conference on Information Systems Design
Methodologies: Improving the Practice, Amsterdam, North-Holland Publishing Co., pp.
87-105 (1986).
19. IConceptual Modelling Language : An Informal Description Report, Institute of Computer
Science, Crete (1987).
20. Theodoulidis C., Wangler B., Loucopoulos P.: Requirements Specification in
TEMPORA. 2nd Nordic Conference on Advanced Information Systems Engineering,
Kista (1990).
21. Malinowski E., Zimányi E.: A conceptual model for temporal data warehouses and its
transformation to the ER and the object-relational models. Elsevier. Data & Knowledge
Engineering, vol. 64, issue 1, January 2008, pp. 101–133 (2008).
22. Kelly Th. J.: Dimensional Data Modeling, Solutions, Sybase,
http://www.olap.it/Articoli/DimensionalDataModeling.pdf accessed August, 2013 (2013).
23. Teki S., Grube M., Griffiths T.D.: A unified model of time perception accounts for
duration-based and beat-based timing mechanisms. Frontiers in Integrative Neuroscience,
January 2012, vol. 5, article 90, pp. 1-7 (2012).
24. Lobato R.S., Ulson R.S., Santana M.J.: A Revised Taxonomy for Time Wrap based
Distributed Synchronization Protocols. Proceedings of the Eighth IEEE International
Symposium on Distributed Simulation and Real-Time Applications (DS-RT 2004),
October 21-23, 2004, pp. 226-229 (2004).
25. Hobbs J. R., Pan F.: An Ontology of Time for the semantic Web. ACM Transactions on
Asian Language Information Processing, vol. 3(1), March 2004, pp. 66-85 (2004).
26. Hobbs J.R., Pan F.: Time Ontology in OWL, W3C Working Draft 27 September 2006,
available at http://www.w3.org/TR/owl-time/ (2006).