- It is a well known fact that information is one of the most important assets for today's enterprises since it is the basis for organizational decisions. However, as information is produced from data, both data and information quality must be assured. Although many researches have proposed technical and managerial solutions to some specific information quality problems, an integrative framework which brings together these kinds of solutions is still lacking. Our proposal consists of a framework for assessing and improving information quality through the concept of Information Management Process (IMP). An IMP is assessed according to an information quality maturity model by using an assessment and improvement methodology. The framework provides a consistent roadway for coordinating efforts and resources to manage information quality with a strategic perspective. As an application example, a study case has been included in the paper.
Aof a data manufacturing process, where data must be
s [
Data is said to be non- ideal due to different and specific
potholes related to special characteristics of data: multiple
sources of data can generate different values; systematic
errors can generate information losses; a large amount of
data can be unmanageable for an application in a
reasonable time; distributed and heterogeneous systems can
generate inconsistent formats, values or definitions, … [
These potholes are all sources of ongoing information
quality problems, like unused data, barriers to data accessibility
or data utilization difficulty. The need to consider data and
information as one of the most important assets for
organizations [
Fortunately, more and more organizations have at last
realized the importance of information quality and they try to
implement some of the frameworks proposed by
researchers [
Unfortunately, there is not an integrative framework that
guides organizations to achieving information quality goals
through management by implementing the
abovementioned concepts [
In an effort to fill this void, we propose an integrative
framework in which information is considered as a product
(what allows the user to take an engineering point of view
of information quality [
Having in mind this perception, entire organizational
information quality can be managed by assessing and im- information quality in organizations. By optimizing the
proving each IMP in organization taking into account that most important and critical IMPs, an organization can reach
several IMP can share several resources. It is true there are a satisfactory information quality state.
several frameworks for assessing and improving software The remainder of this paper is structured as follows: section
processes such as CMM (Capability M [
The main idea of the framework is to use EVAMECAL for assessing and improving an IMP using CALDEA’s lev- 2.2 Definition Level els as reference in the guidance of the optimization of the An IMP is said to be at Definition Level or Defined when it has been defined and planned. This implies identifying all its components and their relationship to the requirements.
To achieve this goal, the following KPAs need to be satisfied: • (IQATM) Information Quality Assurance Team
Management. The aim of this KPA is to form a team
composed of people having direct responsibility for
information and for its integrity. This team will
encourage the entire organization to take on
commitments regarding information quality policies [
Components. For each information quality
component, information quality dimensions from URS-IQ
must be identified, controlled and monitored. It is
necessary to identify from the URS-IQ the dimensions
of quality of information that must be controlled [
An IMP is said to be at Integration Level or Integrated
when besides having been Defined (Definition level has
been achieved), many efforts are made in order to assure
that the IMP is in compliance with organizational
information quality requirements and standards. This implies
standardizing different information quality learned lessons in
order to avoid previous errors and improve future work..
The following KPAs must be satisfied:
• (VV) Information Products and IMP Components
Validation and Verification. Both information
products (obtained as a result of data transformation
process) and IMP components must be verified and
validated to correct defects and/or discord with the
USRIMP, USR-IQ and the organizational information
quality policies.
• (RM) Risk and Poor Information Quality Impact
Management. Authors like [
An IMP is said to be at a Quantitative Management Level
or quantitatively managed when after having been
Integrated (Integration level has been achieved) several
Measurement Plans have been developed and implemented and
measurement procedures have been automated. Therefore,
the main information quality goal of this level is to obtain a
quantitative compliance that IMP performance over a
reasonable time period, remains as consistent as required in
terms of variation and stability through a reliable set of
measurements [
An IMP is said to be at Optimizing Level if when being quantitatively managed the obtained measurements are used to develop a continuous improvement by eliminating defects or by proposing and implementing several improvements. The following two KPAs must be satisfied: • (CADPM) Causal Analysis for Defect Prevention Management. From the study of the measurement results, some typical quality techniques and tools like Statistical Control Process (SPC) or Ishikawa’s diagrams can be applied to detect defects of information quality and identify their root causes. The obtained conclusions must form a basis for a corresponding maintenance process for removing detected defects in affected resources. • (IODM) Innovation and Organizational Development Management. Similarly to the previous KPA, here the results can be used to improve the IMP, in terms of performance, planned time or budget. This is the basis for the idea of continuous improvement.
In the proposed framework, a KPA can be in one of these states: {“Fully Satisfied”, “Satisfied”, “Partially Satisfied” and “Not Satisfied”}. An information quality maturity level is said to be achieved when all contained KPAs are at least “Satisfied”, that is to say, in order to achieve higher levels of CALDEA, the Information Quality Management Team must guide an IMP to the “Satisfied” state of all contained KPA for the lower level. This goal can be achieved by applying the assessment and improvement methodology, which is next described. 3
As previously mentioned, an assessment and improvement methodology is required in order to guide organizations to reach higher information quality maturity levels for each IMP. Basically, the methodology consists of a PDCA cycle. The following is a brief summary of the steps to be taken: 1. 2. 3.
Choose an IMP which needs to be optimized Elaborate a Plan for its assessment.
Execute the assessment plan by conducting the surveys and measuring as required. (see section 3.1 where a set of surveys are described) 4. Analyze the results and elaborate an Improvement
Plan. 5. Study the viability of the Improvement Plan and the solution. 6. Execute the Improvement Plan. 7. Confirm the improvements and obtain and standardize conclusions.
As demonstrated in this document, the numbering for sections upper case Arabic numerals, then upper case Arabic numerals, separated by periods. Initial paragraphs after the section title are not indented. Only the initial, introductory paragraph has a drop cap.
For the assessment process, a set of surveys has been elaborated. This set consists of four different classes of questionnaires, with different goals: A. In order to delimit and characterize the organization, a total of fifteen questions.
B. In order to delimit and characterize the IMP to be assessed, a total of six questions.
C. In order to assess the degree of achievement of each maturity level, several questions organized in different and selective blocks have been developed. The questions are focused on the KPAs, activities, tasks, proposed techniques, tools and practices and required developed products. The idea of organizing the questionnaire in several depth levels is so that questions will be asked from top down and only if necessary. Thus, the block of questions of the first depth level serves to evaluate if KPAs are satisfied or not, avoiding at this depth level other questions which are not important for establishing specific aspects about the accom-plishment of the more specific issues, which are dealt with in lower depth levels. So, if all the answers to the questions of the first depth level differ from “Not Satisfied”, then questions in the level immediately below should be answered, and so on. Altogether one hundred and ninety questions would be answered in the case of all the answers to the questions of the first and the second depth level differing from "Not Satisfied". On the other hand, a maturity level of a given IMP can be achieved a level only if the lower ones have been achieved. Table 1 gathers the number of questions made by each maturity level and each depth level. Due to length restrictions none of the questions are included in this paper. It is important to say that we are working on checking the validity and efficiency of each block and question inside the blocks for each depth level. The answers to these questions must be a number between 0 and 100, in order to quantitatively assess the degree of satisfaction for each task, activity and KPA. Thus, it is possible to set a numeric qualification for each state by calculating a weighted average of the obtained qualification in each one of the KPAs for that level according to a proposed weight given by a critically degree. In the proposal, it has been established that if this qualification is between 0 and 20, the KPA is said to be “Not Satisfied”; if it is between 20 and 60 is said to be “Partially Satisfied”; if it is between 60 and 90 it is said to be “Satisfied”; otherwise it is said to be “Fully Satisfied”. Table 2 shows in the first column the KPA of each level of maturity and in the second one, the degree of criticality of each KPA for that level. The degrees of criticality also serve as a criteria in order to choose which KPAs must be satisfied first in the third step of EVAMECAL. These degrees of criticality are a hypothesis according to the supposed degree of importance. It is a line of future research to determine these weights based on the demands of the different organizations interviewed and to set finely the ranges for qualification.
D. Finally, in order to collate and compare the answers with previous questionnaires, there is a last block of questions with descriptive and textual language.
The surveys are going to be conducted at assessment time.
In order to empirically validate the framework and test its practical applicability and efficiency, it has been applied to different IMPs from several organizations. The following shows the results obtained from applying the framework to a particular organization with proven experience in the information management field. EVAMECAL was first applied by following the steps previously detailed. The results of questionnaires A and B are presented in subsection 4.1 and 4.2; in 4.3 the results of questionnaire C and a list of proposed improvements are presented. Since finishing this paper, the organization has continued working on the improvement plan, although several subgoals have already been achieved.
The main activity of the company is software development with a solid knowledge of and training in software quality standards and Software Engineering (all the developments are carried out by following one of the most important and widely used national software development methodologies). The company has obtained an ISO 9000 certification. Their offered services are consulting, development, training courses, technical attendance, sale of licenses, database and data warehouse administration, system planning projects and migration to an important commercial DBMS (e.g. Oracle of which they are certified partners). With a total of eighty-nine employees, it is the eighteen of the department of systems who organize the resources of computing support for the rest of the departments.
Among all the IMPs, the framework was applied to that related to the Training Management Process, which is a responsibility of the consulting department. The main goal of the IMP in question is to manage data regarding training, which consists of gathering both internal and external requests for training, choosing who are going to be the trainers, determining which resources are going to be used and managing several quality training issues This process is adequately specified and documented in the quality manual of the organization.
Some different forms exist for gathering data about course demands, assignations and quality evaluations of proposed exercises, didactic materials, trainer capability, installations, assistance and used resources.
The organization runs a software application, which is an internally developed tool to manage all previously mentioned data. One of the employees of the consulting department, normally always the same one, is responsible for transcribing data from the forms to the tool and for obtaining the information, which will be given to the adequate person.
All the questions in the surveys were made to the head of the consulting department. In table 3 the main results of these surveys can be found. These results reflect that none of the KPAs belonging to level 2 or higher are at least “Satisfied”. This means that the Definition level is at a “Not Achieved” state. As an example, the conclusions drawn are included: • (IQATM) Information Quality Assurance Team Management. In spite of not properly satisfying this KPA, the organization presents a quality infrastructure that can adequately support information quality. • (IPM) IMP Project Management. This KPA is also not satisfied. • (URM) User Requirements Management. User Requirements have been managed for the training procedures, although information quality requirements have not been taken into account. • (DSTM) Data Sources and Data Targets Management. In the IMP definition, both data sources and data product targets are identified and documented. There are some forms in order to standardize data interchange formats. • (ADMPM) Database or data warehouse Acquisition, development or maintenance Project Management. The database where data is stored is an organizational one adequately modified to support the training managerial software tool. Thus, data and procedure models were modified and extended, although none of the information quality issues were considered. • (IQM) Information Quality Management in IMP Components. This KPA has not been satisfied because information quality is not one of the goals of the IMP. Taking into account the criticality degree of each of the KPAs (see table 2) the following recommendations were proposed in order to satisfy KPAs at Definition level: 1. Create an Information Quality Assurance Team which assumes the responsibility of IMP project management 2. Adequately manage user requirements, both IMP and IQ ones. 3. Identify and define both data sources and data product targets, as well as the data interchange formats. 4. From these requirement specifications, adequately manage the information quality dimensions for each component of the system 5. Modify database or data warehouse to give support to the information quality 6. Plan a project for the development of the IMP. Definition Level Not Achieved (IQATM) Information Quality Assurance Not Satisfied Team Management (URM) User Requirements Management. Not Satisfied (DSTM) Data Sources and Data Targets Partially Satisfied Management. (ADMPM) Database or data warehouse Acquisition, development or maintenance Partially Satisfied Project Management i(nIQIMM)PICnfoomrmpoantieonntsQ. uality Management Not Satisfied Integration Level Not Achieved (CVoVm)poInnefonrtsmVaatiloidnatiPornoadnudctVseraifnicdatioIMn.P Not Satisfied (RM) Risk and Poor Information Quality Not Satisfied Impact Management t(iIoQnSMM)anInagfoermmeanttion Quality Standardiza- Not Satisfied (OIQPM) Organizational Information Not Satisfied Quality Policies Management Cuantitative Management Level Not Achieved (MM) IMP Measurement Management Not Satisfied t(iAoMnMP)anIMagPemMeenats.urement Plan Automa- Not Satisfied Optimizing Level Not Achieved (PCreAvDenPtMio)n MCaaunsaaglemAennatlysis for Defects Not Satisfied (DI OevDeMlop)mInenntoMvaatinoangemanednt.Organizational Not Satisfied
In this paper, an integrative framework for assessing and improving information quality for organizations has been briefly presented, and one experience of applying this framework has been described. The framework consist of two main components: CALDEA, an information quality model based on maturity levels which serves as references when assesing and guidance when improving; both activities, assessement and improvement are supported by the second component, EVAMECAL.
On one hand, the IMP concept with both CALDEA and
EVAMECAL satisfies the four goals required for a good
information quality framework [
On the other hand, the experience of applying the framework to real case studies has allowed both CALDEA and EVAMECAL to be refined and has demonstrated that although it is known that information quality is becoming increasingly important, organizations do not have or do not provide enough time or resources to deal with it. This fact is aggravated by the lack of an information quality enterprise culture.
Although the presented framework is becoming more widely used much work has to be done, beginning with a validation of both models, all the questionnaires of the surveys, and the criticality degrees for each KPA. Another line of work being pursued, consists of choosing (or even developing when they do not exist) some standards, practices, techniques and tools in order to satisfy the majority of the information quality requirements for the majority of the organizations.
This research is part of both CALIPO- supported by Dirección General de Investigación of the Ministerio de Ciencia y Tecnología (TIC2003-07804-C05-03)- and MESSENGER project - supported by Consejería de Ciencia y Tecnología de la Junta de Comunidades de Castilla-La Mancha (PCC-03-0031).