- In the processes and optimization of information integration, such as query processing, query planning and hierarchical structuring of results to the user, we argue that user quality priorities, data inconsistencies and data quality differences among the participating sources have not been fully addressed. We propose the development of a Data Quality Manager (DQM) to establish communication between the process of integration of information, the user and the application, to deal with semantic heterogeneity and data quality. DQM will contain a Reference Model, a Measurement Model, and an Assessment Model to define the quality criteria, the metrics and the assessment methods. DQM will also help in query planning by considering data quality estimations to find the best combination for the execution plan. After query execution, and detection of inconsistent data, data quality might also be used to perform data inconsistency resolution. Integration and ranking of query results using quality criteria defined by the user will be an outcome of this process.
Tbeen widely discussed and researched for a number of he problems of data inconsistency in data integration have years, and a large number of these have been resolved, as described in our own work [35], [36]. However, the combination of these solutions, and the resolution of the remaining issues, still remains an open issue. This has exacerbated as the development of Information Systems, network communications and the World Wide Web, has permitted widespread access to autonomous, distributed and heterogeneous data sources. An increasing number of databases, especially those published on the Web, are becoming available to external users. User requests are converted to queries over several data sources with different data quality, but the quality of the data sources utilised is not a feature of the process.
Integration of schemas on existing databases into a global unified schema is an approach developed over 20 years ago, [4]. However information quality can not be guaranteed after integration, because data quality is dependent on the design of the data and its provenance [31], [5]. Even greater levels of inconsistency exist when data is retrieved from different data sources.
On the other hand, different expectations exist on the quality of the information, depending on the user. A casual user on the Web does not expect complete and precise information [21], but close to his selection condition. A professional user expects accuracy and completeness of the information retrieved in order to make a decision irrespective of the time it could take to retrieve the data, although speed is still likely to be a lesser priority.
User priorities, data inconsistencies and data quality differences among the participating sources have not been fully addressed in the processes and optimizations of information integration, such as query processing, query planning and hierarchical structuring of results to the user. 2
Data integration is the process of extracting and merging data from multiple heterogeneous sources to be loaded into an integrated information resource [4]. Solving structural, syntactical and semantic heterogeneities between source and target data has been a complex problem for data integration for a number of years [28], [4],[35], [36].
One solution to this problem has been developed through the use of a single global database schema that represents the integrated information with mappings from global schema to local schemas, where each query to the global schema is translated to queries to the local databases using these mappings [4]. The use of domain ontology, metadata, transformation rules, user, and system constraints have resolved the majority of the problems of domain mismatch associated with schematic integration and global schematic approaches. However, even when all the mappings, semantic and structure heterogeneity are solved in the global schema, consistency may not have been achieved, because the data provided by the sources may be mutually inconsistent. This problem has remained because it is impossible to capture all the information and avoid null values. At the same time, each autonomous component database deals with its own properties or domain constraints on data, such as accuracy, reliability, availability, timeliness and cost of data access.
Several approaches to solve inconsistency between databases have been implemented: 1. By reconciliation of data, also known as data fusion: different values become just one using a fusion function (i.e. average, highest, and majority), depending on the data semantic [16]. 2. On the basis of individual data properties: associated with each data source (i.e. cost of retrieving data, how recent is the data, level of authority associated with this source, or accuracy and completeness of data). These properties can be specified at different levels: the global schema design level, the query itself or in the profile of the user [2].
Some definitions of data quality criteria, metrics and measurement methods are presented in the following sections.
“High data quality has been defined as data that is fit for use by data consumers and is treated independent of the context in which data is produced and used” [29].
Data quality has been characterized by quality criteria or dimensions such as accuracy, completeness, consistency and timeliness [31], [16], [8], [22], [29], [25] and [20]. However there is no general agreement on data quality dimensions [32], [14].
There has not been a specific differentiation between IQ and DQ, because the terms data and information are often used synonymously. However, Data quality is related to accuracy and integrity and on the other hand, Information Quality is concern with data quality in context, and is related to how the information is produced and interpreted.
A definition of quality dimensions and a framework for analysis of data quality as a research area was first proposed by Richard Wang et al. [32]. An ontologically based approach was developed by Yair Wand et al. [31], this model analyzed data quality based on discrepancies between the representation mapping from real world (RW) to information system (IS) and vice versa, through design and operation activities involved in the construction of an information system as an internal view. A real world system is said to be properly represented if there exists an exhaustive mapping, and no two states in RW are mapped into the same state in IS. Four intrinsic data quality dimensions were identified: complete, unambiguous, meaningful and correct. Additionally mapping problems and data deficiency repairs were suggested. The analysis produced a classification of data quality dimensions as related to the internal or external views. Data Quality measurement method was not addressed (See table 1).
A different classification of data quality dimension was developed by Diane Strong et al. in [29] is based on a dataconsumer perspective. Data quality categories were identified as intrinsic, accessibility, contextual and representational. Data quality measurement method was not addressed. Each category was directly addressed to different data quality dimensions (See table 2).
In Total Data Quality Management (TDQM) [33] the concepts, principles and procedures are presented as a methodology which defines the following life cycle: define, measure, analyze and improve data as essential activities to ensure high quality, managing data as a product. There is no focus on multi-database integration, or data inconsistency detection
TABLE 1 CLASSIFICATION BASED ON INTERNAL OR EXTERNAL VIEW [31]
(design operation)
(use, value)
Accuracy, reliability, timeliness, completeness, currency, consistency, precision System-related: Reliability
Timeliness, relevance, content, importance, sufficiency, usableness, usefulness, clarity, conciseness, freedom of bias, informativeness, level of detail, quantitativeness, scope, interpretability, understandability System-related: Timeliness, flexibility, format, efficiency “Inaccuracy implies that Information System (IS) represents a Real World (RW) state different from the one that should have been represented” “Whether the data available are the true values (correctness, precision accuracy or validity)” “The degree of correctness and precision with which real world data of interest to an application domain are represented in an information system.
Ambiguity: Improper representation: multiple RW states mapped to the same IS state “Ability of an IS to represent every meaningful state of the represented real world system. Thus is not tied to datarelated concepts such as attributes, variables, or values” “The extent to which data is not missing and does not have sufficient breadth and depth for the task at hand” “All values for a certain variable are recorded” “Whether all the data are available” “The degree to which all data relevant to an application domain have been recorded in an information system.” “The IS state may be mapped back into a meaningful state, the correct one” “The extend to which data is correct and reliable” “Whether the data is out of date, An availability of output on time” “The extent to which data is sufficiently up to date for the task at hand” The degree to which the recorded data are up-to-date” “How fast the IS state is updated after the real world system changes.” Age: of data, when first received by the system Delivery time: when data is delivered by the user Input time: When data is received by the system. “Whether the data are up to date, reflecting the most recent values” “The rate of change of the real world.” “Refers to the length of time data remains valid.” “Refers to several aspects of data. In particular, to values of data inconsistency would mean that the representation mapping is one to many. This is not considered a deficiency.” “The extent to which data is presented in the same format” as consistent representation “Often referred as integrity constraints state the proper relationships among different data elements” “The degree to which the data managed in an information system satisfy specified constraints and business rules.”
Wand /Wang Motro/Rakov Gertz Pipino/Wang Ballou Motro Gertz Wand/Wang Pipino/Wang Wand/Wang Pipino/Wang Gertz Wand/Wang Pipino/Wang Motro Wand/Wang Pipino/Wang Wand/Wang Pipino/Wang Motro Gertz
RW/IS states Data values Data model (table, row, attribute, classes) schema column population RW/IS states Currency Volatility Age Delivery time Input time Time RW/IS states Values of data on Integrity constraints Data representation. Physical rep. data Values of data on Integrity constraints Source of data S Accepted stand. A Previ. experience P Time request TR Time delivery TD Time no longer useful TN.
Data path A.
Structure B Path lengths C
1 – ( #incomplete items / #total items) 1 - (# errors /
# total) Max (0, 1 - (# currency /
#volatility)) Age + delivery time – input time Time data invalid - Time start valid 1 – ( #inconsistent / #total consistency checks) Min(A,S,P) Max (0, 1 – (TR – TD /
TR – TN)) Min (A,B,C) or database retrieval solutions.
There are just definitions, and in the best cases, measurement of data quality aspects.
In table 3, the different quality dimension definitions are presented with the relevant factors on each dimension and the proposed metric by author.
Information Quality (IQ) criteria have been classified in an assessment-oriented model by F. Naumann in [20], where for each criterion an assessment method is identified. In this classification the user, the data and the query process are considered as sources of information quality by themselves, (see Table 4.)
The AIM Quality Methodology (AIMQ) [34] is a practical tool for assessing and benchmarking IQ organizations, with three components: PSP/IQ Model which presents a quality dimension classification by product quality and service quality using information consumer perspective, and consolidates the dimensions into four quadrants: sound, dependable, useful, and usable information, these quadrants are relevant to IQ improvement decisions. IQA instrument measures IQ for each IQ dimension. In a pilot study, using questionnaires answered by information collectors, information consumers, and IS professionals in six companies, these measures are average for the four quadrants and the scale used in assessing each item ranged from 0 “not at all” to 10 “completely” and the IQ Gap Analysis Techniques assess the information quality for each of the four quadrants. These gap assessments are the basis for focusing IQ improvement efforts. This methodology uses questionnaires as main measurement method, taking a very pragmatic approach regarding IQ.
In the following section we will present some approaches demonstrating how a data quality model, assessment methods and user priorities, based on the work discussed above, can help in the process of data integration. 3
Database integration is divided by Motro and Rakov [16] in two main problems, intensional and extensional inconsistencies. Intensional are related to resolving the schematic differences between the component databases, this issue is also known as semantic heterogeneity. Extensional inconsistencies are related to reconciling the data differences among the participating databases [16]. Information integration is the process of merging multiple query results into a single response to the user. There are several important areas of related work to consider in the following approaches.
Data integration techniques have been developed by Gertz [8], [9] based on data quality aspects within an object oriented data model, and data quality information stored in metadata. Quality aspects such as timeliness, accuracy and completeness were considered in the process of database integration. The main aspect was the assumption that quality of the data stored at different sites can be different and the quality varies over time. Query language extensions were necessary to support the specification of data quality goals for global queries and thus data integration. In the case of data conflicts between semantically equivalent objects, the object with best data quality must be chosen. If no conflicts exist between objects but their quality level is different, the integrated objects need to be grouped to allow the ranking of the results.
The project MULTIPLEX directed by Motro and Rakov [16], addressed the problem of extensional inconsistencies and a Data Quality Model for Relational Databases. MULTIPLEX was based on accuracy and completeness as quality criteria, this model assigned a quality specification for each instance of a relation, and these quality specifications were calculated by extending the relational algebra. The quality of answers was calculated by the measure of arbitrary queries from the overall quality specification of the database [16]. In the case of multiple sets of records as possible answers to one query, each set of records has an individual quality specification. A voting scheme, using probabilistic arguments, identifies the best set of records to provide a complete and sound answer and ranking of tuples in the answer space. The conflict resolution strategy, and the quality estimates are addressed by the multidatabase designer.
An enhancement of the Multiplex system FUSIONPLEX [2], [3] stores information features or quality criteria scores in metadata, the considered quality dimensions are timestamp, accuracy, availability, clearance and cost of retrieval. Inconsistencies are resolved by data fusion, allowing the user to define data quality estimation on a vector of features weights, performance thresholds and a fusion function at attribute level, as required. This approach reconciles the conflicting values at attribute level using an intermediate result named polyinstance, which contains the inconsistencies. First the polyinstance is divided in polytuples, and using the feature weights and the threshold, members of each polytuple are discarded. Second each polytuple is separated into mono-attribute polytuples using the primary key, assuming that the same value of the primary key between databases refers to the same object but with different data values, and attribute values are discarded based on corresponding feature values. Finally the mono-attribute tuples are joined back together resulting in single tuples.
Information Quality reasoning is defined as the integration of information quality aspects, to the process of planning and optimizing queries against databases and information systems by F. Naumann in [21]. Such aspects are related through the establishment of information quality criteria, assessment methods and measure.
Selection of data sources, and optimization of query planning by considering user priorities has been also addressed in [21] by the definition of a quality model and a quality assessment method under the following assumptions: 1. Query processing: Concerned with efficiently answering a user query to a single or multi database. In this context efficiency means speed. 2. Query planning: Is concerned with finding the best possible answer given some cost or time constraint. Query planning involves regarding many query execution plans across different, autonomous sources that together form the complete result.
In this approach information sources were selected by using Data Envelopment Analysis method (DEA) [6], and the following quality dimensions: understandability, extent, availability, time and price, discarding sources with poor quality before executing the query.
However different sources have different quality scores and they must be fused to determine the best quality result, the quality fusion can be done in two ways 1) applying a fusion function per each quality criteria and find the best combination to query [17] or 2) computing the information quality score using different quality criteria such as availability, price, accuracy, completeness, amount response time for each plan and thus a ranking of the plans using Simple Additive Weighting method (SAW) explained in [11].
The completeness of the query result derived from different sources is approached in [24] considering the number of results (coverage) and the number of attribute values in the result (density). Completeness is calculated as the product between the density and the coverage of the corresponding set of information sources.
In this seminar, it was established that it is essential to first concentrate on developing expressive data quality models, and once such models are in place, develop tools that help users and IT managers to capture and analyze the state of data quality in an information system. [10]. 4
Databases have traditionally been considered to be sources of information that are precise and complete. However the design and implementation of such systems is carried out by human beings, whose are imperfect, so during the whole software life cycle errors occur that are reflected in the quality of both software and information. Furthermore, when these sources of data come from different applications, distributed both physically and logically, these errors multiply. In the field of Information Systems, this shortcoming has been realized and frameworks and models of reference have been developed as standards, such as ISO 15504 [12] and CMMI [1], [7].
Here, the general objective is to establish good practices for software engineering and to be able to talk the same language during software processes, no matter the architecture or implementation methodology. The same challenge need to be taken up in the Data Quality area, based on the following:
1. It is essential to identify a framework that establishes the models corresponding to the criteria of quality, methods of measurement, assessment and improvement, and considers the data quality life cycle.
This framework can be used as good practice during information system development, integration, capture and tracking of changes in data. Tracking changes should offer quality improvement and data cleaning based on a feedback provided by the same information system or a set of recommendations to the information manager, and will help to achieve self regulating systems.
2. This framework might be considered in heterogeneous systems, before, during and after the integration of information.
3. We propose a Data Quality Manager as the mechanism to establish communication between the user, the application and the process of integration of information, to deal with semantic heterogeneity problems, as part of the framework mentioned above (see Figure 1.)
Reference
Model Measurement
Model Quality
Metadata Assessment
Model Data Quality Manager
Selection of data sources
Query Planning Detection and Fusion of Data inconsistencies
Query Integration
Ranking query results
Information Integration Process Fig. 1. Data Quality Manager in the process of information integration.
4. The Data Quality Manager will contain the following elements: • • •
Reference Model: In this model the data quality criteria will be defined depending on data sources, users and application domain.
Measurement Model: This will contain the definition of the metrics to be used to measure data quality, also the definition of a quality metadata (QMD) and the specification of data quality requirements such as user profiles, query language.
Assessment Model: The quality scores definition is essential to establish how the quality indicators are
going to be represented and interpreted.
5. The Data Quality Manager will establish the basis for taking decisions during the identification of data sources in heterogeneous systems, such that: • To classify the sources of data based on certain criteria of quality, depending on the application domain. The scores must be stored in a metadata for every source of data (see Figure 2.) • • Data
Fusion
The use of quality aspects previously stored in the metadata as a whole with the user priorities for the selection of the best sources of information before the execution of the queries, for example if the user prefers those sources of information that are more current with regard to those of major credibility (see Figure 3.)
Help the query planning process by considering data quality estimations to find the best combination for the execution plan (see Figure 4.)
Quality User Priorities QueryA QueryB QueryC QMD
Quality Metadata Definition Assessment of Data
Sources Data Sources Involved in the
Query QMD
Top ranking Query Plan
Fig. 4. Query Planning After query execution, and detection of inconsistent data, data quality might be used to perform data fusion (see Figure 5).
Execute Query Plan QMD
ResultX ResultY
ResultZ Quality user priorities
Data Fusion
Data Inconsistencies
Detection Inconsistent
Query Result
Consistent
Query
Result
Integration of the information sources ranking with the quality criteria estimated by the user (see Figure 6)
ResultJ ResultK
ResultL Consistent
Query Result
Quality user priorities
Query Integration
QMD
Query Result
Ranking
Fig. 6. Ranking of Query Result CONCLUSION
We have shown that, although there has been considerable past work in the resolution of semantic heterogeneity in multi data source systems over a number of years, expressive data quality models and tools to utilise them remain to be developed [10]. The approach developed for Information Quality reasoning [21] provides some mechanisms for data source selection, but does not address many of the data quality factors identified in Table 3. Accordingly, we propose a Data Quality Manager as a framework to deal with data inconsistencies and lack of quality due to different sources; presenting a continuous process of data validation, such as definition of quality criteria, selection of best data sources, ranking of query plan, detection and fusion of data inconsistencies and ranking of query result considering quality of data sources and user expectations. This work is already under way and performance reporting of the tools developed will appear in the next twelve months. [19] F. Naumann and C. Roker, ʺDo Metadata Models meet IQ Requirementsʺ, Proceedings of the International Conference on Information Quality, MIT Cam‐ ACKNOWLEDGEMENT bridge, 1999. [20] F. Naumann and C. Roker C., ʺAssessment Methods for Information This work was supported by financial funding from Consejo Quality Criteriaʺ, Proceedings of the International Conference on Informa‐ Nacional de Ciencia y Tecnologia CONACYT, Mexico. tion Quality (IQ2000), Cambridge, Mass., 2000. [21] F. Naumann, ʺFrom Databases to Information Systems‐Information REFERENCES Quality Makes the Differenceʺ, Proceedings of the International Confer‐ ence on Information Quality (IQ2001), Cambridge, Mass., 2001. [1] D.M. Ahern, A. Clouse, and R. Turner, ”CMMI® Distilled: A Practical [22] F. Naumann, ʺQuality‐Driven Query Answering for Integrated In‐ Introduction to Integrated Process Improvement”, The SEI Series in formation Systemsʺ, Lecture Notes in Computer Sciences LNCS 2261, Software Engineering, Addison Wesley Professional, 2003. Springer Verlag, Heidelberg, 2002. [2] P. Anokhin and A. Motro, “Data Integration: Inconsistency Detection and [23] F. Naumann and M. Haeussler, ʺDeclarative Data Merging with Con‐ Resolution Based on Source Propertiesʺ, Proc. of FMII 2001, 10th International flict Resolutionʺ, Proceedings of the International Conference on Informa‐ Workshop on Foundations of Models for Information Integration. Viterbo, Italy., 2001 tion Quality (IQ2002) Cambridge, Mass., 2002. [3] P. Anokhin and A. Motro, ʺFusionplex: Resolution of Data Inconsis‐ [24] F. Naumann, J. Freytag and U. Lesser, ʺCompleteness of Information tencies in the Integration of Heterogeneous Information Sourcesʺ, Sourcesʺ, Workshop on Data Quality in Cooperative Information Systems Technical Report ISE‐TR‐03‐06, Information and Software Engineering (DQCIS2003), Cambridge, Mass., 2003.
Dept., George Mason Univ., Fairfax, Virginia, 2003. [25] L. Pipino, W.L. Yang and R. Wang, ʺData Quality Assessmentʺ, Com‐ [4] C. Batini, M. Lenzerini and S.B. Navathe ʺA comparative Analysis of munications of the ACM,vol. 44 no. 4e, pp.211‐218, 2002. Methodologies for Database Schema Integrationʺ, ACM Computing [26] [Parsian99] A. Parssian, S. Sumit and V. Jacob, ʺAssessing Data Qual‐ Surveys, vol. 18, no. 4, pp. 323‐364, 1986. ity for Information Productsʺ, Proceeding of the 20th International Con‐ [5] P. Buneman, M. Liberman, C.J. Overton and V. Tannen, “Data Prove‐ ference in Information Systems (ICIS1999), Charlotte, North Carolina nance”, http://www.cis.upenn.edu/~wctan/DataProvenance, [(date in‐ USA, pp. 428‐433, 1999.
[27] E. Pierce, ʺAssessing Data Quality with Control Matricesʺ, Communi‐ formation as accessed by the author citing the references, e.g. 17 Aug. cations of the ACM, vol.47, no. 2, pp.82‐86, 2004. 2004.)] [28] A. Sheth and L. Larson, ʺFederated Database Systems for Managing [6] A. Charnes, W. Cooper, and E. Rhodes. “Measuring the efficiency of Distributed Heterogeneous and Autonomous Databasesʺ, ACM Com‐ decision making units”, European Journal of Operational Research, pp. puting Surveys, vol. 22, no. 3, pp.184‐236, 1990. 429‐444, 1978. [29] D.M. Strong, W.L. Yang and R.Y. Wang, ʺData Quality in Contextʺ, [7] M.B. Chrissis, M. Konrad and S. Shrum “CMMI®: Guidelines for Process Communications of the ACM, vol. 40, no. 5, pp.103‐110, 1997. Integration and Product Improvement”, The SEI Series in Software Engineering, [30] D.M. Strong, W.L. Yang and R.Y. Wang, ʺ10 Potholes in the Road to Addison Wesley Professional, 2003. Information Qualityʺ, Proceedings of IEEE, vol.18, no. 9162, pp.38‐46, [8] M. Gertz and I. Schmitt, ʺData Integration Techniques Based on Data 1997.
Quality Aspectsʺ, 3rd National Workshop on Federal Databases, Magde‐ [31] Y. Wand and R. Wang, ʺAnchoring Data Quality Dimensions in Onto‐ burg, Germany, 1998. logical Foundationsʺ, Communications of the ACM, vol. 39, no. 11, [9] M. Gertz, ʺManaging Data Quality and Integrity in Federated Data‐ pp.86‐95, 1996. basesʺ, Second Annual IFIP TC‐11 WG 11.5 Working Conference on Integ‐ [32] R.Y. Wang, V.C. Storey, and C.P. Firth, ʺA Framework for Analysis of rity and Internal Control in Information Systems. Warrenton, Virginia, Data Quality Research,ʺ IEEE Trans. Knowledge and Data Eng.,1995. Kluwer Academic Publishers, 1998 [33] R. Wang, ʺA Product Perspective on Total Data Quality Manage‐ mentʺ, Communications of the ACM, vol. 41, no. 2, pp.58‐65, 1998. [10] M. Gertz, ʺReport on the Daugstuhl Seminar, Data Quality on the Webʺ, [34] L. Yang, D. Strong and R. Wang, ʺAIMQ: A Methodology for Informa‐
SIGMOD Record, Vol. 33, No. 1, Mar. 2004. tion Quality Assessmentʺ, Information and Management, vol. 40, no. 2, [11] C.L. Hwang and K. Yoon, “Multiple Attribute Decision Making: pp. 133‐146, 2002.
Methods and Applications: a state‐of‐the‐art survey”, Berlin; [35] L.M. MacKinnon, D.H. Marwick, H. Williams, “A Model for Query DecomSpringer‐Verlag. position and Answer Construction in Heterogeneous Database Systems”., [12] ISO/IEC Joint Technical Committee 1 (JTC1), Subcommittee 7 (SC7) Journal of Intelligent Information Systems, 1998. Working Group 10 (WG10) page, there are nine parts of ISO 15504. [36] H. Williams, H.T. El‐Khatib, L.M. MacKinnon, “A framework and 1998. test‐suite for assessing approaches to resolving heterogeneity”, In‐ [13] H. Kon , E. Madrick, and M. Siegel, ʺGood answers from bad dataʺ, Sloan formation and Software Technology, 2000.
WP#3868, 1995. [14] G. Tayi, D. Ballou and Guest Editors, ʺExamining Data Qualityʺ, Communica‐ Pilar Angeles obtained her first degree in computer engineering from the tions of the ACM, vol. 41,no.2, pp.54‐57, 1998. Universidad Nacional Autonoma de Mexico (UNAM), in 1993, [15] U. Leser and F. Naumann, ʺQuery Planning with Information Quality DMiepxloicmoa(IiTnAEMx)pienrt19S9y4s,teDmipslofmroaminTTheeleInmsatittiuctoSyTsetecmnoslofgroicmo IATuAtMonionm1o99d5e, Boundsʺ, Proceedings of the 4th International Conference on Flexible Query An‐ and M.Sc. in Computer Science, regarding Quality in Software Enginnering swering, (FQAS00), Warsaw Poland, 2000. from the UNAM in 2000. Since 1989 she has been working on Technical [16] A. Motro and I. Rakov I, ʺEstimating the Quality of Databasesʺ, Proceedings Support for Databases in Casa de Bolsa Probursa, Nissan Mexicana, of FQAS 98: Third International Conference on Flexible Query Answering Systems, Software AG, Sybase de Mexico and e-Strategy Mexico. Recent research T. Andreasen, H. Christiansen, and H.L. Larsen, ed., pp. 298‐307. Roskilde, iisnteareFsutnsdhearvMe einmcbluedreodf Dthaeta“QQuuaaliltiytyinanSdoHftwetaerreogEenngeinoeuesriDngataMbeaxsiceasn. SAhseDen.mark, Springer‐Verlag, Berlin, Germany, 1998. sociation” (AMCIS). [17] F. Naumann, ʺData Fusion and Data Qualityʺ, Proceedings of the New Tech‐
niques & Technologies for Statistics Seminar. Surrent, Italy 1998. [18] F. Naumann, ʺQuality‐driven Integration of Heterogeneous Informa‐ tion Systemsʺ, Proceedings of the 25th Very Large Data Bases Conference (VLDB99), Edinburgh, Scotland, 1999.
Lachlan M. MacKinnon is Reader in Computer Science, and Director of Postgraduate Study in Computer Science, at Heriot-Watt University. He has a first degree in Computer Science, and a PhD in Intelligent Querying for Heterogeneous Databases. He researches and consults widely in Data, Information and Knowledge Technologies. He is a member of the IEEE, British Computer Society, ACM, AACE, immediate past Chair of the British