Data Inventory Draft Technical Report
|
|  Yüklə 0.75 Mb.
|
2Operational ApproachAlthough conceptually coherent, the method described above has some practical limitations, when implemented. The definition of the outputs purely based on user’s needs may imply the development of applications and data collection campaigns which may not be implemented with the available resources or within the given timeframe. This approach does not take into account other constraints such as the institutional/logistic frameworks and data policy issues that may invalidate the development of some of the applications/models or make impracticable the collection of essential data. To avoid such problems the current work has taken a more pragmatic approach in which the identification of outputs and applications are not independent from the available information and resources. This implies that outputs and system functionalities might be adjusted to “fit” the available data if resources are not sufficient or available to collect information in the form it might be desirable. The example reported in Box 1 explains this concept. T
The information on delivery of relief items such as seeds or agricultural tools is important for evaluating when the beneficiaries are expected to receive the items. An estimate of the delivery time could be made using information on past purchases if the dates of order and delivery of the items were available. While the order and payment dates are recorded by the FAO Procurement Service, the delivery date is not available. Using the payment date instead of the delivery date, it is possible to estimate with a good approximation the delivery time. This might require the application to take into account the nature of the data and to apply correction factors accordingly. he process of looking at the outputs as functions of the available information and resources results in: (a) the exclusion of certain outputs because not feasible, (b) the adaptation of database applications in accordance to the types of data and (c) the identification of data to be collected because considered essential and therefore irreplaceable. A characteristic of this approach is its cyclicality where outputs, applications and data are reviewed in rotation to redefine the database functionality, as it is described in Figure 2. 
Figure 2 In this way, the feasibility of the identified modules – and consequently the generation of the associated outputs - is ensured by the existence and availability of the necessary information and resources which are verified during the decisional process. 3Issues Regarding the Accessibility of FAO DatabasesO Box 2 In the previous sections the words “database” and “information system” have been used in an almost interchangeable way. It is now due to clarify that as a consequence of the dynamic nature of the information framework in FAO and in other related data holding agencies/institutions, the accessibility to emergency-related information involves the collection of a variety of objects, of which data is only one of the object types. For example, interfaces able to extract information from other systems and make them available to the emergency officers in the appropriate format are objects as important as data. The typology of structures with the above characteristics is typical of an information system intended as a collection of data, application and reports. It would be therefore more appropriate to refer to such structure as an Emergency Information System (EIS) rather than an Emergency Database. ne of the driving criteria used in this initial phase, was to maximize the use of data already available in FAO HQs and to establish links with in-house operational information systems. Data and systems should be seen as a unique body since the issue of data integration in FAO is strictly related to the integration of the various systems managing and distributing the information. As a direct consequence of this, the Emergency Information System (EIS) might rather be an entry point for accessing data from other systems than an actual data repository (see Box 2). In the majority of the cases, FAO information is published through the Internet and/or Intranet and is provided with interfaces that enable the selection of information using keywords or other searching criteria. While it is certainly an advantage that datasets in FAO are produced, maintained and distributed by the responsible divisions, the accessibility to this information within the EIS has to be thoroughly evaluated. Replicating software functionalities, which were already developed for other applications, might not be appropriate and in principle the emergency database could exploit to the maximum the existing capacities. Although a fast and economic solution could, then, be to use the existing software applications to access the data from the various divisions (i.e. establishing simple links to the websites), these applications are not specifically designed for supporting emergency operations and therefore are not provided with the necessary tools and functions. Such solution would reduce the user’s ability to analyze and output data the way it is required for his/her purpose decreasing the efficiency of the EIS. A viable solution could be to develop interfaces (between the EIS and the other FAO applications) able to extract datasets and indicators relevant for emergency applications. This solution has been investigated and discussed with various potential data providers such as AFIS (FPMIS), SDRN (GeoNetwork), ESCG (GIEWS) and AFSP (FAO ORACLE Procurement Services). The idea had positive response from the contacted officers who confirmed its feasibility from the technical point of view and their availability to exchange information with the EIS. However, once objectives and data/software requirement are clearly defined, proposals for the development of the identified applications have to be prepared to evaluate timeframes and financial resources. In this respect, some of the issues to be taken into account are (a) structure, formats and content of the data to evaluate efforts for data conversion and (b) potential drawbacks, with particular regard to the efficiency of data-flow to and from the EIS. Speed of connection and downloading time might, in fact, affect the efficiency of the EIS and could represent a constraint that is likely to increase if the EIS is forced to connect to different databases every time an operation is performed. In order to reduce this problem it would be convenient to limit online connections only for those datasets that are frequently updated such as ARTEMIS, FPMIS or ORACLE Procurement. Other databases, especially if not very demanding in terms of disk space such as FAOSTAT, could be transferred on regular basis to the Emergency Data Server. This would also be an inexpensive solution which saves costs for the developing interfaces for accessing the information. Moreover, independently from the technical solution adopted to access and publish the information, clear and unambiguous statements indicating the use and distribution of the information should be formulated and agreed between the PAIA-REHAB and the other partners. |