An audit of records management was conducted from September to November 2000 to assess DFAIT's management of paper and electronic records. The audit reviewed operations at four Missions, three Headquarters bureaux and the groups in the Information Management and Technology Bureau (SXD) which provide functional support to the Department. The audit scope excluded electronic records maintained in DFAIT's functional business systems such as WIN Exports and PeopleSoft.
The audit observed that work units have generally developed and maintain their own processes to manage their paper and electronic records. As a result, they are using a variety of practices to meet their information needs. Paper records are kept by individual staff or in the unit's filing cabinets. Although there are exceptions, these records are generally transferred to the central records system or disposed of only when the units run out of filing space. In the exceptions, the units transfer records to SXD's Central Processing Units (CPU's) on a regular basis. The practices at Missions are similar, although very little is sent to their records offices (formerly known as registries).
There is a greater variety of places to store electronic records than paper records so, not surprisingly, there is less consistency in how they are managed. Electronic records can be found on: Corporate Automated Text Storage (CATS); Classified Corporate Automated Text Storage (CCATS); shared network drives (I:\ drives); personal network drives (H:\ drives); personal and organizational e-mail folders; Intranet and Internet servers; Outlook Public Folders; as well as on individual diskettes and compact disks.
CATS is the repository for non-classified corporate electronic documents. Documents are put into CATS by individuals who either send them there as email attachments or "cc" CATS on their outgoing email messages. The degree to which individuals send items to CATS varies from person to person and depends on staff remembering to send all key documents to CATS. The majority of staff interviewed agreed that they do not always send electronic records to CATS, but some added that the omission was usually not intentional. The absence of a prompt to remind people to "cc" CATS was cited as one reason, as was uncertainty as to what actually should or should not be sent there.
When asked of their experience retrieving documents from CATS, most individuals replied that they have not tried to do so directly. The majority of those that did reported only limited success. This could be due to the limitations of CATS not being explained to users - for example, documents with certain key words in the text are protected and will not appear in searches conducted by general users. The Headquarters staff who used the SXD's Local Information Service Offices (LISO's) to conduct searches for them had more success.
The effective handover of electronic records upon rotation seldom happens. New officers usually have access to their predecessors' paper files, but few reported that these are useful. Some units used the shared network drives (I:\ drives) extensively for critical program information. In these cases, the impact of staff rotation is lessened as the predecessors' files remain available, subject to how well they are organized.
The result of these practices is that for the most part, individual work units have generally found ways to meet their current information needs. However, longer term corporate memory that is kept in records is being lost. As well, full corporate compliance with the requirements of the Access to Information Act and the Management of Government Information Holdings Policy is doubtful.
Some of the individuals interviewed expressed concern about their own practices, but did not know where to go for help. The answer at Headquarters should be to SXD's LISO units and at Missions to their Systems Administrators. Program staff have little knowledge that the LISO's role is to provide expert advice and guidance to program units on information and records. Although the Client Services Division (SXC) indicates that the LISO's have begun carrying out this role, there was no sign of such activity in the three bureaux reviewed. LISO involvement in these bureaux was limited to conducting searches.
LISO searches cover not only the electronic records in CATS and CCATS, but the paper records in the corporate repositories. These repositories are maintained by the Corporate Processing Units (CPU). There is an extensive backlog of documents, approximately 200,000, waiting to be classified and/or indexed by CPU staff. This backlog has existed for years; there were 50,000 records reported in the backlog during a 1985 audit. The audit team observed a general activity level in these offices that was less than we would have anticipated, given the backlog. This lack of urgency is reflected in workload throughput that is well below CPU standards, which themselves appear to the audit team to be very low. In general, there is a hands-off approach to managing the LISO's and CPU's.
There are no comprehensive policy and guidelines setting out how program units, LISO's and CPU's should be managing records. Various guidelines, practices and tips are on the SXD intranet site, but almost all program staff interviewed did not know they exist. A comprehensive policy needs to be communicated which will provide a vision of how DFAIT manages records. It should present an accountability framework and performance objectives while addressing where there is flexibility and where there is not. The need for a clear policy was also mentioned in some of the studies recently conducted by consultants and SXD staff in this area.
A number of recommendations are summarized below in the Summary of Recommendations and detailed in the report, addressing
Two options are also presented for consideration. The first is to establish a records management "swat" team that would go from division to division, quickly reviewing business practices and developing and implementing tailored records management solutions. The second is to explore with senior management the option of effectively replacing the current systems and procedures and implementing a leading edge solution.
The Department is actively addressing the findings and recommendations of the Audit which was undertaken at the request of the Information Management and Technology Bureau. Action has already been taken to reduce backlogs, improve communications and training and to evolve our information management policies. The Department is also working closely with other government departments and with central agencies that are focussed on a renewed approach to information management in the Government.
The two options for consideration proposed by the Audit are complimentary and are being pursued in tandem. A model office initiative aimed at improving the departmental information management practices division by division has been launched. The Department has also initiated an Integrated Document Management Project which will replace the current departmental systems. Together these projects will be used to further the goal of improving information management practices across the Department.
Below is a summary of the audit's recommendations:
|1||SXD||Update the Trade records classification system used at Missions to incorporate new industries.|
|2||SXD||Develop and implement a communications strategy to promulgate a corporate perspective on records management to all staff, with appropriate references to the "how to" information currently on the SXD web site.|
|3||SXD||Provide additional information on the CATS search web-page to inform users of the limitations on the searches they can conduct.|
|4||SXD||Consider instituting a mandatory prompt within the email program which would require staff to select whether or not each outgoing message should be sent to CATS.|
|5||SXD||Update the current CATS filing and retrieval capability or consider replacing the system to provide a viable, and corporate-wide, research tool.|
|6||SXD||Initiate a proactive approach by records staff to assist programs in establishing, using and monitoring their records holdings, regardless of where they are stored.|
|7||SXD||Develop a comprehensive "made-in-DFAIT" records management policy which takes into account the operational requirements of Headquarters and Missions as well as corporate records requirements.|
|8||SXD||Develop and implement a communications strategy to promulgate the policy to ensure that staff understand and implement the policy.|
|9||CFSI||Build on the basics learned in SIGNET 2000 training by developing a "layered learning" approach to increasing staff competency in electronic program uses.|
|10||CFSI||Develop this layered approach through an inventory of current technology irritants and wish lists obtained from staff to be trained.|
|11||CFSI||Recognize the value of staff time taken from operations to conduct these training sessions and adopt a "minimal time, maximum applicability" approach.|
|12||CFSI and SXD||Consider making at least some training in Records Management mandatory.|
|13||SXC||Conduct education sessions with staff in their working environment, learning how various records management practices, and tools, would apply to their type of operation.|
|14||SXC||Ensure that LISO staff receive the appropriate training themselves on how to provide the necessary training to staff.|
|15||SXC||Implement a timely, routine series of reminders for staff, on how to conduct certain key practices to improve records management.|
|16||SXD||Implement a proactive approach to facilitating organizational records management by SXD, through the LISO's at HQ and the Records Office and SA operations at Missions.|
|17||SXD||Implement continuous interaction between records management specialists and their clients. This means that records staff must understand the operational information requirements of their clients, and provide advice as to how best to meet the client's, and corporate, information needs.|
|18||SXD||Consider establishing a small task force of information management experts to systematically begin reviewing and improving the practices of Headquarters units.|
|19||SXI||Develop a plan with on-going targets with the goal of clearing the paper records backlog.|
|20||SXD||Regularly monitor the progress of SXI in clearing the backlog.|
|21||SXI||Determine and apply the minimal indexing throughput requirements that would allow the Department to cost-effectively store and retrieve its paper documents.|
|22||SXI||Revisit the allocation of resource requirements and the staffing needs to bring the service resources to an acceptable level.|
|23||SXI||Conduct periodic monitoring of throughput results, providing feedback to unit staff and managers, taking corrective action where required.|
|24||SXIS||Conduct periodic document retrieval tests to assess the processes followed, the throughput times experienced, and to determine where difficulties may indicate classifying and indexing problems.|
|25||SXCA||Rotate the assignments of LISO staff to allow for knowledgeable back-up to cover illnesses, vacations and training.|
|26||SXD||Revisit the philosophy of the current LISO/CPU structure, compare intentions to actual experience, and determine which structure will best achieve the intended results.|
|27||SXD||Strengthen the linkage between LISO and CPU staff, either by organization realignments or by improving the liaison that currently exists.|
|28||SXD||Institute a program of on-going liaison between LISO/CPU staff and program units to ensure that departmental staff are aware of, and abide by, records retention guidelines.|
|29||SXD||Conduct periodic visits by LISO/CPU staff, or where available, Mission Records Clerks, to the various program units to assess records retention situations and to take appropriate action.|
|30||SXD||Periodically review the records in the CPU backlogs to see if they can be disposed of.|
|31||SXC and SXI||Invoke "hands on" management within LISO's and CPU's, including the establishment of realistic objectives, service standards and throughput targets.|
|32||SXC and SXI||Conduct regular monitoring of achievements to date, assess reasons for any negative variances, and take corrective action where necessary.|
|33||SXCA||Foster effective client/LISO working relationships as soon as possible to understand the client's business requirements and assist management and staff to conduct more effective and efficient records management practices.|
|34||CIO (SXD)||Obtain the views of senior management on benefits of implementing state-of-the-art records / information management systems and practices. Then, if warranted prepare a thorough business case.|
An audit of records management was conducted from September to November 2000. The purpose of this audit was to assess DFAIT's management of both paper and electronic records. The key objectives were to determine if DFAIT was meeting both its program and legislative requirements, and if it had struck the right balance in doing so.
In DFAIT, program officers and individuals all have responsibility for records management. The Information Management and Technology Bureau (SXD) has functional responsibility for records management and supports departmental staff in meeting these records management responsibilities. The audit reviewed the operations of the central records management groups in SXD which provide functional direction and support to Headquarters units and Missions.
To assess the decentralized records management practices used by individuals throughout the Department, practices at a sample of four Missions and three Headquarters bureaux were reviewed.
The Missions (Vienna, Lisbon, Copenhagen and The Hague) provided a cross-section of Mission records management scenarios. They varied in size and other factors.
|Mission||Approximate # of Staff||Head of Administration||Systems Administrator||Records Maintenance|
|Vienna||100||CB MCO||CB||Records Clerk|
|Hague||43||CB MCO||LE||Records Clerk|
|Lisbon||25||CB MCO||LE||HOM SCY|
|Copenhagen||16||LE MAO||LE||HOM SCY|
CB MCO: Canada-based Mission Consular Officer
LE MAO: Locally-engaged Mission Administration Officer
HOM SCY: Head of Mission secretary
Excluded from the scope of the audit were electronic records maintained in DFAIT's functional business systems. As such, the audit team did not review consular information in the COSMOS system, interactions with the Canadian and foreign business community in WIN Exports, extensive records on export and import permit transactions in IPPS and ECS, personnel records in PeopleSoft nor financial records in IMS.
What is Records Management?
Departmental information is a resource which must be managed, as the human, financial, material and property resources of the Department are managed.
A record is information in any format, created or received and maintained by the Department, a work group or an individual in the transaction of business or the conduct of affairs and kept as evidence of that activity. Records are required by the Department to control, support or document the delivery of programs, to carry out operations, to make decisions, to account for the activities of the Department or to provide evidence in legal actions.
Records management is the function that enables the Department to meet its legal obligations to manage both paper and electronic recorded information sources. Included in the corporate record-keeping system are the policies and guidelines for the creation, identification, classification, retrieval, receipt and transmission, storage and protection, disposition and preservation and sharing of information and records. A record-keeping system includes the space, equipment and personnel required to administer the records management function.
The responsibility to meet the records/document management obligations rests with all employees, not just records management personnel. Every departmental employee who creates, receives or is the custodian of information that controls, supports or documents the delivery of programs, is responsible and accountable for managing that information and ensuring it is included in the corporate record-keeping system. All employees of the Department have a responsibility to create and manage the records necessary to document their official activities. All the information that an employee of the federal government collects or creates in the conduct of business is the property of the Government of Canada.
The responsibility to establish and maintain the corporate record-keeping system is that of the Department's Chief Information Officer (CIO). The CIO is also the Director General, Information Management and Technology Bureau, and is responsible for the overall management of the Department's information management and technology program. In support of this mandate, records and information management staff collect, organize, store, maintain and retrieve records and provide assistance to departmental program clients.
The Department has a legal requirement to collect and maintain government records. Government of Canada policy, as outlined in the Management of Government Information Holdings Policy, and the National Archives of Canada Act sections 5 and 6, requires all federal departments to:
More information on the government requirements related to records management are provided in Appendix B.
All programs of the Department and "owners" of large collections of information are obligated to manage information as a corporate resource in accordance with government legislation and policies. Most program areas at Headquarters and Missions keep isolated islands of electronic and paper records. Electronic records are stored on shared drives (I:\ drive), occasionally in Public Folders, and in personal Outlook Mail (email) accounts and non-shared drives (H:\ drive). Many paper records are stored in "divisional" file cabinets managed by individual units, and in personal file cabinets.
A number of initiatives have recently been undertaken within SXD to investigate and address departmental records and information management issues. Within the past year, we are aware of five such studies:
Information Management Action Plan
Status: in progress
Assessment of the Indexing of Paper Documents
Process Review of the Management of the Corporate Record
Status: In progress
Public Folders - Review and Renewal
Departmental Active Files Consolidation Project
With this number of studies, that audit team hopes that management will have the information it needs to quickly decide and take action on the audit's recommendations.
During the audit, the records management practices of a number of units from various programs were reviewed. This included units both at HQ and at four Missions in Europe. Through a review of these units' practices, the Audit team believes it has been able to understand the general practices in the Department as a whole. While there will always be particular units for whom all of these observations will not apply, the audit team is satisfied that there was enough consistency in the inconsistency of practices to draw overall conclusions.
Practices in each of the units reviewed brought out the different approaches taken by staff, and programs, to managing their records, both the paper documents they handled and the electronic documents they received and transmitted to others. These differences have arisen as units and individuals have essentially been left to fend for themselves in developing their practices, particularly in that area of electronic records.
Another common feature is the impact on records management when staff rotate out of positions, whether at HQ or at Missions. This fact of life at DFAIT has a profound effect on the preservation, or loss, of paper and electronic records. While nothing can replace the depth of knowledge a person will have of issues after being in a position for a period of time, the lack of reliable and complete records due to poor handover practices, is exacerbating the problem.
Most rotational officers interviewed at Headquarters said they had neither the time nor the need to review the paper records left for them in their new offices. Some left the records as is while others conducted a cleanup of these documents, purging unnecessary documents and forwarding others to the Corporate Processing Units (CPU). Any new hard copy records that officers created or received are either filed according to their own record-keeping needs in their offices, destroyed when considered appropriate, or sent to the CPU.
Generally, Mission officers involved in political, economic, UN or other agency programs opted for limited reviews of previous hard copy documents left to them by predecessors. We did find exceptions to this observation, some of which are noted in Appendix A. Clearly though, the majority of records left by predecessors were found to be of little use to current officers. The audit team did not assess whether these documents should have been sent to Mission records offices.
Added to the volume of DFAIT generated records which Mission officers needed to manage was the onslaught in some situations of externally generated documents from UN or single purpose agencies. One only had to visit their offices to witness this phenomenon of stockpiling paper records.
Most officers, even those working in similar groups, had different paper records management practices. As well, there appeared to be little records management assistance provided to these programs by Mission Records Clerks, where they existed, other than filing documents that were sent to the records office (formerly known as the Registry).
The role to be played by Mission records offices bears scrutiny. Of the four Missions visited, two operate records offices with dedicated CBS (CR's) for this function. These staff deal only with records sent to the central records office and they do not assist program officers in managing their paper records. The result is cabinets full of handover documents from previous officers. Trade sections generally operate their own records offices, with no help from the CBS records staff as an LES handles this function. Annual destruction of official records was conducted recently at two of the four Missions visited while one Mission had not purged old records for many years.
Records in the records offices of all four Missions were organized using the 1973 records classification system, although one Mission was in the process of cleaning out files and converting the remainder to the new RICS system which is used at Headquarters. Consistency in records classification and management therefore does not exist between Headquarters and Missions; however there is no obvious need to integrate these systems. Several staff members did say that the Trade filing system needs to be updated to incorporate new industries and technologies.
Recommendation to SXD:
1. Update the Trade records classification system used at Missions to incorporate new industries.
1. The records classification system used at HQ (RICS) was developed to accommodate the requirements of the Trade program. We propose to encourage missions to adopt a simplified form of RICS and to support missions that chose to follow that direction.
Electronic records can be found in many places, with CATS being the official corporate storage location. Other options available are the I:\ drive, personal H:\ drives, personal and organizational email folders and Public Folders. The use of these four options varied from Mission to Mission and within units at Headquarters. These options applied to the unclassified records, as CCATS was the automatic electronic storage source for classified records on the C4 system.
Documents are put into CATS by individuals who either send them there as e-mail attachments or "cc" CATS on their outgoing email messages. The degree to which individuals sent items to CATS varied from person to person. It depends on staff consciously remembering to send all key documents to CATS when authored or received from an outside source by them, or to ensure that upon receipt from other DFAIT sources the document had already been copied to CATS by the initiator. The majority of staff interviewed agreed that they do not always send electronic records to CATS, but some added that the omission was not usually intentional. Views on the value of adding a prompt to remind people to "cc" CATS varied, with some welcoming the idea and others being opposed to it.
With the number of electronic storage options available for staff, it became clear that there is no consistency in electronic records management. The use of the I:\ drive at the Missions visited was normally for Mission administrative and general information purposes, it was not being used to store critical program documents. While the I drive at some Missions was well organized, at others the I:\ drive was accurately described as "a mess," and "anarchy."
At HQ, some Bureaux use the I:\ drive as a critical source of program information. At one geographic Bureau, the structure and usage has been in place for a few years. It contains current fact sheets, briefing notes and the history of officer correspondence considered key to the Bureau's programs. As a result, it is the primary source of local corporate memory for the Bureau. Even when staff moved on, the new officers could access their predecessors' critical correspondence, briefs and other program information using this I:\ drive. The I:\ drive also became the eventual source for Bureau intranet web-site material allowing Missions, and others, access to Bureau information.
One corporate liaison division uses the I:\ drive to store all current Q's & A's developed by originating divisions. Their I:\ drive is structured to identify such records by month, by Minister and by document reference number. Each month the I:\ drive's contents are transferred to CD ROMs and to the division's Public Folders. The responsibility to copy the Q's & A's to CATS is considered to be that of the originating divisions, and not the corporate liaison division.
The management of the H:\ drive and email was at the users' discretion at all locations reviewed. Some developed structured folders to manage their electronic mail and authored documents. Others filed most records without any structure. Some deleted records on a regular basis, depending on the currency of the document or the relevance to their current projects, topics or events. Still others had not cleaned out or organized their electronic records storage sources for months, which usually mirrored their paper records habits. Appendix A provides some specific examples.
Rotationality has a significant impact on electronic records management. In the majority of situations reviewed, outgoing officers did not provide the incoming officer with a copy of the electronic records they had been maintaining. There were a few exceptions, however, which allowed new officers access to electronic records on diskettes or CD ROMs. Where I:\ drives or Public Folders are in use to maintain divisional records memory, the impact of rotating staff on corporate memory appeared to be reduced.
The use of Public Folders within the Outlook program is very limited. Many staff were not aware of its existence, and those who did had virtually no training in its use. Three exceptions were noted in the units reviewed:
Details on the use of Public Folders for ten of the units reviewed are found in Appendix A.
Impact of Current Practices - Division, Mission and Individual Level
The impact of the situations described earlier varies from office to office, and from person to person. It appears that most, but not all, operational groups are adequately managing their records to meet their own immediate needs. However, there is little view of the "big picture." Over the years, they and their predecessors have developed ad hoc procedures which have become the de facto standards for the area. Adapted to local conditions and the way their particular unit works, these informal "ways and means" of getting their job done works for them, but only to a point.
Most groups can find the documents they want, when they want, although the efficiency of this process is imperfect and often time-consuming. The same applies to individuals who manage their own records directly. This type of "stovepipe" situation, where individuals or units manage "their own" records, typically leads to information isolation. Almost by definition, the information base these employees are working with is limited, although this is not apparent on a day-to-day basis, at least to individuals working within the unit. They have the information they need, generally, and are not aware of the wealth of information that could be available to them, or others, if the Department's records were properly indexed and universally accessible.
Without being aware of these various information sources, we believe that there is a great deal of duplication of effort being undertaken across various units. Often it has driven the multiple 'cc's found on many email, where many of the recipients are included on a 'just-in-case' basis. Each of these individuals now has to delete, print or file their copy of the message. Other examples are documents received from external sources which are copied to Missions and various Headquarters interested units, which leads to multiple copies of the same record in various offices as opposed to being on a central reference drive. A Headquarters geographical units may maintain an I:\ drive system of briefing notes, but as their Missions do not have access to this drive, each interested Mission must also maintain their own related records system.
There is evidence in some areas reviewed that sharing information, in paper or electronic form, is a key operational process. Where paper documents are the primary media, copies are retained and originals sent to CPU's, Mission records offices or work unit files to ensure that the corporate memory is maintained together with ease of local access.
However, there are information shortfalls. Data available indicates that less than one half of one percent of all unclassified electronic traffic is sent to CATS. What percentage should be referred is difficult to assess. If the information is not in CATS (or any other shared system), employees see no benefit in bothering with the system, so new material is not entered, and the cycle continues.
Most individuals replied that they have not tried to search for documents in CATS directly. Of those that did, most reported only limited success. This could be due to the limitations of CATS not being explained to users - for example:
Unless the existence of a particular memo or letter is known and is not retrieved during a search, the search may be deemed a success, without actually being complete. This situation will adversely affect the broad-based, subject-related research questions more than a search for a particular document. With a reasonable subject search, given a reasonable amount of time, it is our opinion that only partial success is possible the vast majority of the time.
Recommendations for SXD:
2. Develop and implement a communications strategy to promulgate a corporate perspective on records management to all staff, with appropriate references to the "how to" information currently on the SXD web site.
3. Provide additional information on the CATS search web-page to inform users of the limitations on the searches they can conduct.
4. Consider instituting a mandatory prompt within the email program which would require staff to select whether or not each outgoing message should be sent to CATS.
2. A communications strategy is being implemented which includes encouraging employees to seek advice from our intranet documents.
3. Further information will be added to the web-page and employees will received a separate communication explaining search techniques and limitations.
4. A pilot project to test this concept will proceed shortly with a few HQ units. Deployment of such a feature to all users will depend on the results of the pilot project.
Impact of Current Practices - Departmental Level
The Department has to meet the requirements of legislation which pertains to the management, preservation and sharing of records . A list of pertinent legislation is attached as Appendix B. As things now stand, the Department is not fully compliant with many of these legislative requirements.
In particular, we doubt the ability of the Department to respond completely and promptly to requests made under the Access to Information Act. The current process is for the Access to Information and Privacy Protection Division (DCP) to request all relevant documents from the applicable units and the LISO. Based on our observations regarding filing procedures in various divisions, we are doubtful that all the responses from these units are complete.
Also in doubt is the ability of the Department to meet the requirements of the National Archives of Canada Act, especially where it pertains to the preservation of records, which includes electronic records. Our observations on paper and electronic records management at Missions and HQ units support this assessment.
The Department is also expected to efficiently manage and exercise custodianship of its overall resources. When information resources are managed in a haphazard or inefficient way, the productivity of human resources suffers. If staff spend significantly more time looking for a single, accessible, authoritative copy of a document than they have to, that leads directly to less time spent on the Department's primary mandates. This was confirmed in a recent survey of Political Officers:
For current issues, the existing systems and processes may be adequate. As noted, employees and workgroups tend to be able to retrieve the information they need to deal with a specific current situation. What appears difficult, however, is dealing effectively with longer-term issues, issues which transcend a number of years and, perhaps, a number of different staff.
One anecdote that was passed to the audit team related to negotiations with another country that took place in the early 1990's on a particular topic. When the topic recently came up again, no record of the substance of these original discussions could be found by DFAIT (or the foreign government for that matter). As a result, both parties agreed not to spend more time looking for the documents and started over.
Recommendations to SXD:
5. Update the current CATS filing and retrieval capability or consider replacing the system to provide a viable, and corporate-wide, research tool.
6. Initiate a proactive approach by records staff to assist programs in establishing, using and monitoring their records holdings, regardless of where they are stored.
5. The current Information Management Action Plan project will make recommendations on our medium-term IM environment. A corporate-wide search capability is among the requirements in the draft report.
6. Present proactive assistance will be expanded. To begin, process improvement projects will be undertaken with three bureaux.
What is Missing?
What is missing, in general, from the mosaic of services and service levels offered, is a "corporate" perspective. There is no consistent set of standards or service levels applied across the Department. The SXD web-site contains a variety of guidelines, best practices, etc., but this of itself does not constitute corporate policy or direction on records management against which staff can be held accountable. In almost all cases, the staff interviewed were not aware that this information existed. One officer summed up the situation by stating that "I am not being assessed for my records management practices. It is not what I am paid to do."
The overall integration of information, the Department-wide "corporate memory," does not exist, except in bits and pieces. This affects not only its ability to meet legislative and regulatory mandates, but its efficiency at meeting operational goals. Just because most items can be found, eventually, does not mean that the systems are adequate.
Recommendations to SXD:
7. Develop a comprehensive "made-in-DFAIT" records management policy which takes into account the operational requirements of Headquarters and Missions as well as corporate records requirements.
8. Develop and implement a communications strategy to promulgate the policy to ensure that staff understand and implement the policy.
7. As a first step to articulating clear policies, guiding principles for Departmental information management are now under review. Policy writing will follow shortly.
8. Making policies known figures prominently in our communications strategy.
Training - Program staff
To date, training for DFAIT staff in records management has been limited, consisting of general information management courses and the original SIGNET 2000 training. There has been no follow-up training on the electronic tools and no records management training program has been prepared.
The presence of various guidelines, tips, suggestions, and best practices found at various layers within the SXD web site does not constitute a training plan. The audit found no program staff that knew of the existence of these documents on the web site, many of which are excellent and, which if adopted by staff, would greatly improve existing practices.
Both rotational and non-rotational staff require repeated reminders. Without periodic reminders to staff of the existence of these "tips and guidelines", the majority of staff will ignore what is currently available. The approach to date has not been a good communication vehicle to make staff aware of what they can do to facilitate records management practices.
Recommendations for CFSI:
9. Build on the basics learned in SIGNET 2000 training by developing a "layered learning" approach to increasing staff competency in electronic program uses.
10. Develop this layered approach through an inventory of current technology irritants and wish lists obtained from staff to be trained.
11. Recognize the value of staff time taken from operations to conduct these training sessions and adopt a "minimal time, maximum applicability" approach.
9. CFSI has an ongoing layered learning program for IT knowledge and skills. Modules and workshops on particular aspects of information management (i.e. Public Folders, CATS, I Drive Management) exist. Further training opportunities on the broader issue of records management would be added once policies, procedures, roles and responsibilities for functional and user communities etc. are clarified. Modules would be developed upon completion of a needs analysis, which could be done quite directly. Distinctions would be drawn between Information Management at HQ and at Missions.
10. See above. CFSI's approach is to do "needs analyses" of communities targeted for training, rather than developing of irritants.
11. CFSI organizes training on the premise that staff time is precious. Training is designed using the shortest possible time necessary to attain stated objectives. This would be applied to info management training. CFSI would design a light but effective set of options for training on information management. We would ensure that training modules would not be intrusive. A key dimension to this would be to graft such a training effort onto an effective communication strategy. The CFSI Virtual Campus is available for web-based training modules, both self-pace and instructor-led. This would be useful for training on the basic features of information management that might also be applicable to missions. Different training approaches of increasing complexity would be arranged for different target groups. A well-defined system of information management could well lead to decentralized accountabilities for feeding it with information.
Recommendation for CFSI/SXD:
12. Consider making at least some training in Records Management mandatory.
12. Mandatory training may not be necessary once a system has been defined. Were mandatory training to be seen as essential, it would require an unmistakably clear order from the Executive Committee or the Deputies and would have to have ongoing senior management intervention to make it work.
SXD has worked with CFSI to ensure that some aspects of information management are included in training courses for new employees.
Recommendations for SXC:
13. Conduct education sessions with staff in their working environment, learning how various records management practices, and tools, would apply to their type of operation.
14. Ensure that LISO staff receive the appropriate training themselves on how to provide the necessary training to staff.
15. Implement a timely, routine series of reminders for staff, on how to conduct certain key practices to improve records management.
13. Plans for IM communication include development of presentations and handouts to be used in such instruction sessions. Experience gained in work with 3 selected bureaux will be applied to development of communications tools for all Departmental units.
14. LISO staff have received training in delivery of presentations. They are receiving further training to expand their coaching skills. A specific training plan for each employee is being drawn up.
15. Plans for IM communication include regular reminders about IM good practices.
Hands-on support for Program units
The role that LISO units should be playing in actively assisting Program units with their information management practices is clearly set out on the SXD intranet web-site:
The LISO provides expert advice and guidance on the organization, identification, filing, retention and disposal of information and records, including drive management. They design and create work group structures for clients within Outlook Public folders, provide a research and records retrieval service for clients and coach employees on searching and retrieving records in CATS. The LISO promotes good records and information management practices within the Department.
However, to date very little progress has been made in delivering on this mandate. Staffing and training were identified as the primary causes of this delay and SXC indicates that the LISO's are now actively carrying out this role. While it is hoped that this situation will be resolved shortly and the LISO's will effectively get this program going, past history cautions against over-optimism. One option to help get this activity underway would be to establish a small task force of two or three people to quickly jump start the process, with the full LISO team moving in when they are up to speed.
This small task force could start with one Headquarters division and work with them for a few days or a week. They would first understand current information management practices and then determine how best to use the department's electronic and paper systems to support that particular unit's business needs. This solution would be presented to the unit. If it is accepted it would then be set-up and supported with desk training for the unit's staff. The task force team members could be made up of individuals who are experts in organization and filing of paper records, as well as use of the applicable electronic tools. After a few units have been done, new team members could be rotated onto the team to share knowledge and perhaps allow for additional teams to be established, until all Headquarters units had been covered. An approach for Missions could also be developed, based on the lessons learned.
It is recognized that there are some drawbacks to this approach. For one, the task force would use resources that would otherwise be bringing the LISO's up to speed. However, in the audit team's opinion, the records management situation in the client community has reached the point where something should be done quickly to start helping them.
Recommendations for SXD:
16. Implement a proactive approach to facilitating organizational records management by SXD, through the LISO's at HQ and the Records Office and SA operations at Missions.
17. Implement continuous interaction between records management specialists and their clients. This means that records staff must understand the operational information requirements of their clients, and provide advice as to how best to meet the client's, and corporate, information needs.
18. Consider establishing a small task force of information management experts to systematically begin reviewing and improving the practices of Headquarters units.
16. LISO's are engaged with HQ units now and experience from the project with three bureaux will provide the foundation for wider initiatives both in HQ and at Missions.
17. Periodic participation by LISO personnel in divisional staff meetings has proven effective and further interaction by both LISO and CPU personnel with their clients will be encouraged.
18. The project of working with three bureaux on business improvement will build a corps of advisors for wider initiatives both in HQ and abroad.
1. Political Officer Study, September 12, 2000
There are a number of corporate based groups and systems involved in the processing, maintenance and retrieval of corporate records. They include the Central Processing Units (CPU's) and the Local Information Service Offices (LISO's), both under the direction of SXD. The electronic corporate systems are Corporate Automated Text Storage (CATS) and Classified Corporate Automated Text Storage (CCATS). There are a number of "island" units of paper records offices not under the control of SXD, while other islands operate under an arrangement with SXD, but managed by the local unit.
Storage of Records
Records can be found in any of the following places :
For any particular search, the LISO officer may have to look in the four "corporate" information storage sites (CATS, CCATS, paper subject file in the CPU and the CPU backlog) before finding the requested material, if specific documents are requested. If a more general request is made, all four must be searched.
The onus is on the initiating organizations to ensure that corporate records are transmitted, either electronically or paper based, to corporate storage locations, on a timely basis. Also, the guiding principle of not sending paper versions if the document has been sent to CATS should be emphasized, as this unnecessarily increases the workload of the CPU's and LISO's, as well as the operating units.
The Department has no discrete Records Management policy. This leaves the records management organization virtually powerless to require operating units to follow what departmental standards do exist.
Such a policy is the cornerstone of any records management program. It describes what records must be kept, who is responsible for their management, how they are to be disposed of, and also why these "rules" exist. In general, the policy is the departmental tool which explicitly gives the administration the power to enforce the legislation pertaining to records.
In general, the records management group (or equivalent) is tasked with creating standards, standard procedures, standard tools, and an education and training program for all staff which will enable them to meet their responsibilities under the legislation and policy.
Document Backlogs in CPU's
There is a backlog of approximately 200,000 documents across all the CPU's under SXD control. This information has typically been received in the last 5 years and has not been classified or indexed as yet. Most of it has been organized according to either date, subject or geographically (e.g. by country), depending on the source bureau.
The backlog has become an accepted fact of life in the CPU's and has its own file cabinets, stretches of open shelving, and its own organization schemes at each location.
SXI has estimated that approximately half the current resources, operating for a full year, would be required to clean up the backlog. The ease of retrieving backlogged documents depends on how the documents are organized at each location. While retrieval may ultimately result, the time required to do so is the issue. SXI has recently put a plan in place to clear this backlog. Two individuals have been added on a trial basis to see if hiring outside help is a viable option.
Recommendation for SXI :
19. Develop a plan with on-going targets with the goal of clearing the paper records backlog.
SXI Response :
19. Simplification of the processing of paper documents has been instituted. Supplementary personnel will be retained on a contract basis. The IM communications plan will try to shift clients from paper to using electronic storage facilities. Moving emphasis away from labour-intensive paper handling to exploiting the potential of electronic storage is seen as the solution to the chronic backlog problem.
Recommendation for SXD :
20. Regularly monitor the progress of SXI in clearing the backlog.
SXD Response :
20. In addressing the backlog and in overall improvement of IM activity, measurement of results will be required.
Workload Standards versus Reality
Standards for indexing and filing documents in the CPU's exist :
|Team Leader||SI-3||Index and classify 125 documents per month|
|Documentalist||SI-2||Index and classify 175 documents per month|
|IM Operations Technicians||CR-3||Classify 140 documents per day|
|IM Operations Technicians||CR-3||2.6 meters of filing per month|
However, these standards are not being met. The latest SXD internal statistics provided to us from 1997 state that :
With this level of performance, even if it has improved since 1997, it is small wonder that there is a huge backlog of unclassified and un-indexed documents. Indexing and classifying an average of 5 to 7 documents a day, even if there are other duties to perform, is well below what should be able to be accomplished in such a unit.
One of the reasons for this poor productivity may be the "over-indexing" of documents. Over-indexing is a practice of spending excessive time analyzing the content of each document and identifying more key words than are required prior to classification and indexing. Indexing practices were recently studied by consultants on behalf of SXD and they found:
"it is clear from the assessment that the value of the current indexing function may not justify the significant resources required to undertake that function at the present level."(1)
The consultants recommended :
"Assign resources within the CPU to reflect where the greatest value is achieved (allowing all records to be retrievable in a timely manner) vs. where the required level of effort is greatest (indexing paper records)."1
The audit team concurs with this assessment and recommendation.
In addition to the problem with indexing, both the CPU's and LISO's are currently operating with numerous vacant positions. At the time of the audit a competition was underway to staff 9 of the 21 SI-02 positions in the CPU's.
Recommendations for SXI :
21. Determine and apply the minimal indexing throughput requirements that would allow the Department to cost-effectively store and retrieve its paper documents.
22. Revisit the allocation of resource requirements and the staffing needs to bring the service resources to an acceptable level.
23. Conduct periodic monitoring of throughput results, providing feedback to unit staff and managers, taking corrective action where required.
SXI Response :
21. Indexing of paper documents has been reduced. Further adjustment will be driven by measurement of results. Transition of the Department from a paper orientation to a greater use of electronic storage will also be significant in making resource allocation decisions.
22. Vacant positions are being filled to address properly the task at hand. Resource requirements and resource allocation are under re-examination.
23. Improved measurement of results in information management is required. Collection of statistics has been reinstituted and some positive results are already evident. A broader plan for IM measurement will be developed.
Test Retrieval of Documents
A test was conducted by the audit team for the retrieval of 15 specific documents. These documents were selected from either CATS, CPU paper files or from the CPU paper backlog. The LISO's had no significant problems finding 12 of the 15 documents selected. Of the remaining three, only one was "un-findable." The times required to conduct the actual searches ranged from 2 minutes to 1 hour, once the appropriate LISO staff were available. However, due to illness of one individual, it took a number of days to begin searching for 9 of the 15 items. This is because individual LISO staff are relative specialists in their client areas. As such, users must rely on them being available because, as the audit team experienced, the searches do not begin without them.
The LISO staff indicated that in a regular search, with interaction with the client, unlike in our "hands-off" test, the two other documents could have been described more fully with a back-and-forth between the LISO and the requestor. We accept this statement and are satisfied with their ability to find specific documents.
Recommendation for SXIS :
24. Conduct periodic document retrieval tests to assess the processes followed, the throughput times experienced, and to determine where difficulties may indicate classifying and indexing problems.
SXIS Response :
24. Retrieval performance measurement will be included in improved measurement of IM results.
Recommendation for SXCA :
25. Rotate the assignments of LISO staff to allow for knowledgeable back-up to cover illnesses, vacations and training.
SXCA Response :
25. Backup in both LISO and CPO functions will be improved to ensure an appropriate standard of service to clients. Rotation is one approach which will be considered.
Organization of Records Management Units
The Bureau Information Control Office (BICO) concept, which was in effect a combined CPU and LISO office, existed in the Department until 1999. Current experience with CPU's and LISO's suggests that the BICO model may have been appropriate after all. Separating the people who research the information from the people who receive, classify and index it is creating problems. LISO staff are unaware of the types of documents being received by the CPU's, how they are being indexed, and where they are stored. It is only upon receipt of a request that they have a need to review the information caches in the CPU and elsewhere.
Another impact of this separation is reduced operational efficiency. Smaller units are affected more severely by staff turnover, vacations and illness. Bigger units, with more staff, would be better able to re-allocate staff as necessary to meet operational necessities and to cope with major issues such as the backlog.
The audit team recognizes that the current organization model was the result of years of study and deliberation. It is fairly new and the units are still not fully staffed or trained. However, should measurable improvements in performance not be seen in the next few months, serious consideration should be given to reverting to the previous BICO model.
Recommendations for SXD :
26. Revisit the philosophy of the current LISO/CPU structure, compare intentions to actual experience, and determine which structure will best achieve the intended results.
27. Strengthen the linkage between LISO and CPU staff, either by organization realignments or by improving the liaison that currently exists.
SXD Response :
26. Achievement of optimal results will drive decisions about structure. Changes in the IM environment (move from paper; new IM tools and policies) will also be significant in the ongoing review of the structure.
27. Enhanced liaison initiatives will be undertaken and more formal arrangements for alignment of the activities of the two groups will be explored.
Retention of Records
Some records are being kept too long, although the scope of this problem is hard to quantify. Transitory records are being kept in the official system well past their effective date and it is not apparent that the disposition schedules agreed to by DFAIT and the National Archivist are applied in a standard way.
Keeping information too long exposes the Department to legal ramifications: if a court orders that all information be turned over, that means all available information must be provided, not just that information which was legally required to be kept. This could be problematic from a content point of view or from a productivity point of view in that it requires more work to gather information than should be necessary.
Recommendations for SXD :
28. Institute a program of on-going liaison between LISO/CPU staff and program units to ensure that departmental staff are aware of, and abide by, records retention guidelines.
29. Conduct periodic visits by LISO/CPU staff, or where available, Mission Records Clerks, to the various program units to assess records retention situations and to take appropriate action.
30. Periodically review the records in the CPU backlogs to see if they can be disposed of.
SXD Response :
28. LISO/CPU personnel will be prime participants in IM communications initiatives.
29. LISO's are visiting operational units in HQ. Communications initiatives will give all employees an improved understanding of retention policies. Liaison with operational units will include review of retention practices.
30. Disposition of records caught in the backlog is incorporated into standard procedures and this will be accelerated as additional resources are applied to the backlog.
Ideally, there should be one location for staff to search for electronic records. Obviously, there are security considerations that need to be taken into account, but that does not obviate the fact that it is currently difficult to find all associated electronic information.
For paper records, a well-organized records office which holds all information, current and 'fairly recent', is essential to well-serve the needs of the clients. Where this type of office is not practical, such as where staff need quick access to subject files, clients should continue to maintain their own records, but still under the "virtual control" of the departmental records authority - SXD. In other words, the Department should know that the records exist, where they are, and be able to obtain them quickly.
Technically, there is no such thing as a "working file", from a records management standpoint, if the "working file" contains the original or only copy of a document. If a document is created in the course of doing business, it is a record. More pragmatically, early drafts and rough notes that have not been sent to anyone else, can be deemed "transitory" and may be destroyed, but that is virtually the only exception. Though this rule is not generally popular, it is essential to good information management: capture all information at creation or receipt. An exact copy of material which is captured elsewhere in the Department's "corporate memory" may be destroyed at the user's convenience.
Management of LISO's and CPU's
The LISO's and CPU's are currently operating in a "hands-off" manner, with little direct management. Our time spent in these units found the level of activity to be less than we would have anticipated, given the backlog . There is no initiative to meet service standards and, especially, to eliminate the backlog of unclassified and un-indexed material. The Team Leaders of these units are operating in a vacuum, with little guidance or scrutiny from their managers. The audit team believes that this should be addressed as soon as possible. Another study is not required to begin actively managing these units.
Recommendations for SXC and SXI :
31. Invoke "hands on" management within LISO's and CPU's, including the establishment of realistic objectives, service standards and throughput targets.
32. Conduct regular monitoring of achievements to date, assess reasons for any negative variances, and take corrective action where necessary.
SXC and SXI Response :
31. Measures to make management of the LISOs and CPUs more effective are a priority for the Bureau.
32. Measurement of results will be a central theme in action to improve management.
Training - LISO and CPU staff
Until recently, there had been little training provided to LISO and CPU staff that pertains to their job responsibilities. It is essential that all these staff understand what information management is, why it is important, and what benefits it can provide to their clientele. If they don't know, or are unable to communicate this message, the clients can hardly be expected to know. The LISO staff in particular need to become the "salespersons" for the Information Management program, encouraging and mentoring all the staff in their portfolio to manage their information properly. Unless they take the initiative and become part of the management team for the operational divisions, their relevance will continue to diminish. It is hoped that the current training initiative underway for CPU and LISO staff will address this.
Recommendation for SXCA :
33. Foster effective client/LISO working relationships as soon as possible to understand the client's business requirements and assist management and staff to conduct more effective and efficient records management practices.
SXCA Response :
33. Both through regular contact (LISO participation in divisional meetings) and special projects (working on process improvement with three selected bureaux), the objective is to understand business needs and respond with appropriate assistance.
C. The Long-term Solution : Start over again?
Many of the recommendations in this report are in support of incremental changes that should improve the current situation. However, the audit team believes that a more comprehensive change to the Department's information management practices that, in effect, starts anew should be considered. The premise of such a solution would be that there is a place for every record -- paper and electronic -- in the corporate system and that staff would store all records in the right place the first time. At any given time, the location of all records would be known.
Employees should be able to access a user-friendly database to query the existence and location of records they need to consult. With effective standards in place and a properly trained and informed work force working in a cohesive fashion, each user should be able to benefit, not only from the records he or she may have participated in creating and storing, but from all other records created by his/her colleagues throughout DFAIT, assuming the security requirements are met, of course. Furthermore, this would be irrespective of the medium these records are in -either paper or electronic.
Behind this support for users, would be an infrastructure set up to meet their needs and expectations:
On a practical basis, staff would have to know into which "folder" every paper and electronic document they handle should be filed. Therefore each folder in the filing cabinet in their office, their unit's shared file area and electronic filing system would be identified in the corporate record-keeping system. There would be no need to send paper or electronic documents to a separate "corporate records unit" as their local and unit's shared areas would be available corporately.
It is recognized that such a solution would require a significant investment in systems, training and likely additional on-going resources. One of the comments that the audit team heard most often was that staff do not have time for records management. However, a comprehensive plan such as this would realistically require more up-front time on the part of all staff as they have to think about each record as they handle it and decide how to properly handle it. In addition, as each new file folder is created, it would have to be added to the corporate list of files. However, it is unclear whether or not additional resources to support program units would be required. The increased time spent up-front handling each record may be off-set by eliminating the current inefficiencies in retrieving information. Without additional resources to support program units, any major new records scheme would likely be doomed to failure.
Would the benefits of such a system outweigh the costs? Perhaps, but not necessarily. As the majority of the units we reviewed were able to meet most of their current needs, any significant benefits to them of a new approach would not accrue for some time. Five or ten years from now, when most of the current staff are gone, these unit may be faced with situations where it would be helpful or necessary to know what is happening now. The other area of benefit will be corporate users, such as the ATIP office and the Legal Affairs who would benefit from such systems now.
The audit team's opinion is that only the most senior level of management would be in a position to decide if it is worthwhile for DFAIT to attempt to become a leading edge records / information management organization. While the Chief Information Officer could determine the costs of moving in this direction, the value of the benefits in terms of improved delivery of the mandate should be assessed and evaluated by Executive Committee.
DFAIT currently has a study underway in which the Department is looking at where it should go in this area and will identify some steps to get there. The audit team believes that there are at least two options to determining where to go: incremental changes or a more radical overhaul.
Recommendation for the CIO (SXD) :
34. Obtain the views of senior management on benefits of implementing state-of-the-art records / information management systems and practices. Then, if warranted prepare a thorough business case.
CIO (SXD) Response :
34. The current Information Management Action Plan project will describe a target state for information management in the Department and steps to achieve that state. Without prejudicing the final results of the study, interim reports indicate that both cultural change and technological advances in IM will be recommended. Those recommendations would lead to development of a fully-reasoned business case for senior management.
The purpose of this appendix is to demonstrate, by way of example, the variety of records management challenges that are faced by units in the Department, and the practices they have established to address them. Below are summaries of the records practices in 10 units, in both Missions and at Headquarters, that were reviewed.
In addition, there are many islands of paper records management activity that are wholly outside of SXD's area of service including major non-SXD operated file registries. Examples of these islands are:
The Department is bound by the following acts, legislation and policies that relate to the management of information. Relevant sections of the acts are detailed below, i.e. those pertaining to information and records management.
The National Archives of Canada Act states that no record under the control of a government institution, or any ministerial record, shall be destroyed or disposed of without the consent of the National Archivist of Canada.
This also means that government institutions that transfer control of original records outside the federal government also require the consent of the National Archivist.
The act also gives the National Archives of Canada the responsibility to undertake the custody, care and control of the records of government institutions and ministerial records that, in the opinion of the National Archivist, are of historic or archival importance.
The Access to Information Act provides an enforceable right of access to government information, in accordance with the principle "that government information should be available to the public." The act also restricts access to certain kinds of information, such as advice and recommendations to the government and Cabinet confidences, until twenty years after their creation, but does not require that such information be maintained for the twenty year period.
This act seeks to render the government more accountable to the electorate and facilitates informed public participation in the formulation of public policy. It seeks to ensure fairness in government decision-making and permits the airing and reconciliation of divergent views across the country.
The Access to Information Act requires that government produce information created through the use of computer hardware and software and expertise normally used by the government, so long as doing so would not unreasonably interfere with the operations of the institution. This relates to situations where a computer is in constant use producing scheduled outputs and is not available for other purposes.
Recently, a new subsection of the act was added that provides for sanctions against any person who improperly destroys or falsifies government information in an attempt to deny right of access to information. Employees who contravene this subsection are guilty of: (1) an indictable offence and are liable to a prison term of two years and a fine not exceeding $10,000 or both; or (2) a summary conviction offence and are liable for a term of six months or a fine not exceeding $5000 or both. This section does not apply to errors made in good faith without the attempt to deceive.
The Privacy Act extends the present laws of Canada that protect the privacy of individuals with respect to personal information held by a government institution and allows individuals a right of access to their personal information collected by the government. As well, another purpose of the Privacy Act is to ensure that government only collects information that is relevant to its programs.
Personal information used by a government institution for an administrative purpose shall be retained by that government institution for at least two years following the last time the personal information was used for an administrative purpose unless the individual consents to its disposal; or where a request for access to the information was received, until such time as the individual had the opportunity to exercise all their rights under the act.
Section 8 of the Charter of Rights and Freedoms regulates government searches, regardless of the context, for searching or monitoring computers. This charter applies to the government in its role as an employer, in addition to its other roles. As well, the Supreme Court has ruled that individuals can have a reasonable expectation of privacy in their offices.
The Management of Government Information Holdings Policy requires government institutions to manage their records throughout their life-cycle and account for them in a corporate inventory of information holdings. The Department is required to:
The Government Security Policy identifies requirements to ensure that all classified or designated information of the federal government is safeguarded appropriately.
The basic information and instructions in chapters 30 through 38 of the 1986 Manual of Correspondence and Communications are still valid for the management of records and information at Headquarters. Chapter 39 deals with records management at Missions. SXI plans to revise these chapters at some point to address the departmental electronic workplace.
In conjunction with these chapters, the following documents are available to employees outlining departmental procedures and best practices to manage records and information. They can currently be accessed on the Intranet from the "http://intranet.lbp/department/sxd/services/RecordsIM-en.asp" page on the SXD site.
1 Assessment of the Indexing of Paper Documents, September 7, 2000