This book “discusses knowledge curation and reuse, within the context of organisations in a digital world who are struggling with the pace of change, and an increasingly fragmented world of work. With increasingly complex, interconnected supply chains, and distributed businesses, the need to bring trusted knowledge into the hands of front-line practitioners has never been more pressing.”

“The term ‘trusted knowledge’ is not intended to imply some immutable truth on a topic, even within an organisational context, but knowledge whose provenance is clear and can be challenged and refined over time. We use the word ‘truth’ as a shorthand for this current yet evolving body of knowledge.”

“I am addressing this book especially to the management and influencers in any organisation who recognise that knowledge management has become an issue that cannot be solved by yet another technological silver bullet, but requires a fundamental rethink at all levels (cultural, managerial, and in terms of systems). I hope that everyone who believes that knowledge is a core asset for organisations, especially in a digital world, will read this book.”

“Business processes within an organisation do something to ‘entities’, in their most general sense. For example: they run a ‘project’; create a ‘drug’; make a ‘loan’. There are many such processes and their associated entities within an organisation.”

“A key idea in the approach to be championed here is not to introduce artificial ‘hierarchies’ of processes. Instead we ‘follow the work’ (as in ‘follow the money’), and ignore the artificial and often arbitrary departmental boundaries. If a loan application moves across several departmental boundaries, then in hierarchical language we would see a loan dossier moving up and down departmental hierarchies.”, but

“In our approach these boundaries are ignored entirely and instead we show a flat process, with no hierarchies at all. The hand-offs between different ‘roles’ (which just happen to be distributed in different departments) are shown in all their simplicity.”

“The former (hierarchical) way of modelling a process is not very revealing and is really intended for an IT audience not a business one. It has the additional weakness of needing to be re-worked every time a departmental restructuring occurs. Whereas the latter (flat) way of modelling the process has immediate resonance with a business audience in my experience and is resilient to restructuring; because it reflects the essence of the process.”

“While the core data-centric transactional systems that generate income are usually in fair shape (or at least functionally adequate), those that deal with the less tangible arena which ‘knowledge’ occupies are often highly fragmented and reduced to barely more than a jumble of digital filing cabinets - whether situated on the corporate network or in the ‘cloud’ makes little difference. Therefore, in most organisations, we see a wholly inadequate platform for the capture, curation and dissemination of knowledge.”

“Google and Facebook do not have to worry about a company’s intellectual property rights, or customer confidentiality, or the long-term preservation of knowledge, because they are really focused on empowering the digital citizen here and now, and empowering businesses wanting to sell products and services to them.”

“A typical business has quite different motivations. It will have chinese walls separating different functions - required for reasons of good governance or client confidentiality”

“The precision and fidelity required to ensure that the right information is at the finger-tips of the right role, at the right time, is something that is key to an efficient and well run business, and something that a Google search on a digital dustbin will not and cannot deliver.”

“This book is ultimately concerned with helping experts working on any collaborative enterprise with the difficult judgements they make, pertinent to their area of expertise. They often base these judgements on work that is either not openly published … and place that knowledge within the specific context of the moment.”

“… whatever your views on AI, this book is certainly not trying to dismiss it.”

“Knowledge sharing in organisations should not be simply mediated through ‘search’, even sophisticated AI-based search. Far superior is the knowledge embodied in synthesised resources such as practice guides. These are informed by experience from multiple projects, and are carefully curated distillations of knowledge.”

“A document in the digital world addresses the same fundamental need of documents in the paper world.”

but needs to be reimagined.

“Documents are created by people for a purpose, to communicate or substantiate something that is intended to be meaningful, for a specific or general class of people.”

“For a particular ’document object’ we can expect it to possess a number of ‘functional characteristics’, including some or all of the following:

• Has a Title;

• Has Ownership;

• Has a Unique Identifier;

• Has Meta-Data;

• Has Intellectual Property Rights;

• Is Versionable;

• Has a Life-cycle;

• Has Renditions;

• Is associated with a Process;

• Has Role-Based Access Controls;

• Can have Links to multiple locations in information architecture;

• Has Retention rules;

• Has an Audit Trail.”

The book shows how to make this a reality, for a wide range of complexity of document; anything from a memo, to a legal bundle, to an interactive graphical novel, to the design for an airliner.

“The life-cycle of data - from transaction-oriented data-centric systems towards knowledge-centric document oriented systems - is something that all organisations do, albeit unconsciously. It tends to be the result of evolved business practices and many linkages between departments and functions. Some of the links are automated but many will be done manually, such as, cut and pasting a graphical view of some data and inserting into a document; this causes a breakdown in the ability to trace the source of the data.”

“One can judge if access controls have been properly implemented in an organisation, by looking at how much use is made of attachments in emails. If it is high, then it is almost certainly proof that the approach to access control in information systems is poor. If it worked, people would be sending links to content, knowing that the recipient who needs access will gain access via that link, rather than a copy (via an attachment).”

“In an organisation, people change roles, or leave, yet business processes continue unabated. While some processes, like an online purchase, are fleeting, most processes are more extended in duration; in some cases, such as a new drug approval or major construction project, they can extend for years. If I am the author of a bridge design and I need some senior engineers in the department to review the design, do I explicitly list their names - Mary Brunel, Joe Heavside, and Tim Janes - to facilitate their review of a document? What if Tim retires halfway through the project and Susan replaces him? What if there are many documents in several folders that need to have the changes made to access? Won’t this involve a mammoth retrospective change to access control lists throughout the system everytime there is the slightest change in personnel?”

“For these kinds of reasons, the good practice in information management is to avoid attaching explicit names to document roles (such as reviewer) wherever possible, and instead to define logical groups that will perform these roles, and then to attach the names to these groups..”

The book shows how to systematically apply simple but coherent access control strategies.

“A need exists for people who are dedicated to language, and make careful decisions about its usage, within the business context. Linguists, taxonomists and philosophers will often be better candidates to consider these questions than computer scientists because they have the skills required to forensically analyse the use of language, and to develop a consistent set of terms for use in categorizing documents.”

“We cannot think about controlled vocabularies in isolation from metadata. In simple terms, think of a classification scheme as being a set of named fields that are to be associated with documents, and controlled vocabularies as the source of terms that will be applied to the fields. It is only when these values are applied to a specific document - and the conjunction of the field name and its value, Name and Richard Erskine - that they can be referred to as metadata. People will often loosely refer to a controlled vocabulary as ‘metadata’ but this is wrong; only in its application does the terminology become metadata. Taxonomies can be viewed as providing one form of controlled vocabulary. The terms within a taxonomy can be used to populate a metadata field. Where there are multiple taxonomies relevant to different fields, then each taxonomy can be applied in turn to populate the metadata.”

“Rather than expecting the software engineer or systems architect to take on multiple possible disciplines that otherwise do not have a home in the organisation (digital archivist, illustrator, taxonomist, anthropologist, records manager, etc.), additional competencies need to be identified and elevated in their status, building a team with the breadth of skills needed to work towards a vision, and really make it stick.”.

“This is a huge opportunity to diversify the community charged with making a ‘knowledge platform’ a reality. Those with backgrounds in the arts and humanities, can join those from engineering and science in building the competencies required. Let us welcome back disciplines like librarians and records managers, discarded by organisation in the drive to go digital. They can offer important skills, albeit refreshed and reframed in the digitally empowered organisation. We can all learn from each other, and debate how best to translate the vision into reality.”

So IT has an important role, but within a much richer and diversified delivery approach, where business needs to demonstrate leadership and insight.