Taxonomic Process Project Overview


Collection process Linear Project planning.gif
1. Planning

Collection process v3 .gif
2. Collecting
pbv4.jpg
3. Processing Specimens

herbarium-compactus-bronwyn-2008-300.jpg
4. Curation

Research v21.gif
5. Research


2981030929 25b926a616.jpg
6. Publishing

curator large.jpg
7. Maintenance

Taxonomic Processes

Ingredients

Stages (including underlying process and challenges)

  1. Planning
  2. Collecting
  3. Processing of specimens
  4. Curation
    • Curation is relevant to many components and products of the taxonomic process. The Curations page provides links to analyses of these components and products.
  5. Research
  6. Publishing
  7. Maintenance

Project Overview

General Comments: " Beating back the bush. "

The practice of taxonomy is Recursive and Non-linear. It includes processes which are easily comprehended (e.g. collection, accession, & curation) and more contemplative activities which are hard to specify and difficult to document.

Our tactic is to proceed from the easy to the hard through a number of iterations. We hope to circumscribe what is hard to describe. We trust that subsequent consideration (and broader consultations) will tell us more. Along the way, we will collect information about where and how Informatics might facilitate the practice of taxonomy.

Why Recursive?

At its heart, the processes of (biological) taxonomy (see definition) are driven by research programmes into the identification of Biological Taxa (e.g. species) and their relationships. This incorporates processes easily imagined by anyone familiar with other disciplines with an interest in collections management and categorisation. Collections are made primarily to facilitate research. Valued material is held long-term for further research and so large collections arise. Conversely, the identity of taxa, their distinguishing characteristics and, even more so, the underlying hypotheses of taxonomy are informed by what has been collected in the past. Thus, the concepts underlying the organisation of large taxonomic collections and the practice of taxonomy are constantly under challenge and review, which in turn impacts upon the day-to-day operations of collections curation.

Why Nonlinear?

Why? Because the underlying processes are
  1. recursive and/or
  2. include non-linear sub-stages (this is the case with most intellectual endeavours but especially so in a discipline which relies so much on inductive reasoning)
  3. include processes with variable (& not easily predicted) time frames (e.g. peer review, loans )
Time taken to come to produce a result is hard to predict. It is (for instance) not related to the number of specimens at hand nor taxa under revision.

Strategy

The practice of taxonomy varies subtly with discipline and indeed with taxa. Host institutions also exert significant influence on the practice of taxonomy.

It is also influenced by between variation in the resources, practices and agendas of host The information at hand varies across the different stages with practice and discipline. Methods vary with traditions, materials, media, lab practices and goals. Goals vary with discipline, phyla and at fine levels between taxa.

In seeking opportunities for process driven enhancements and technologically driven enhancements, we need to describe
  • what is the same
  • what is particular
for each discipline or institution and then to compare and cross reference
  • terminologies
  • methods
  • challenges.
Heuristics. (probably overworking the metaphor but) The following three approaches are being used to circumscribe the Taxonomic Process:-
Approaching from above (Defining the Field)

We have developed a series of flow charts and process models. These are intended as foils for collecting responses from practitioners. In their first incarnations, they describe the taxonomic process at a high level of abstraction. In no way are they meant to be authoritative descriptions of what taxonomists do. They undoubtedly include stages and processes which are skipped and elsewhere omit or gloss over stages which are essential. We hope to use them as catalysts to help elicit responses from practitioners in the field as we gather information and, hopefully, some good case studies.

You will find a key to the flowcharts below. The meaning of elements has been kept vague and ill-defined on purpose. As our understanding improves, we will revise our description of taxonomic processes. Eventually we hope to have a formal description (as appropriate and where possible) of taxonomic processes in UML or BPL.

Approaching from Below (Defining the elements)

We already know a lot about some of the specific challenges faced by taxonomists in their work (see OtherPreviousProjects? and OtherCurrentProjects?). We are also gathering more information formally (in workshops) and informally (in interviews). Collating what is already to hand gives a skeleton on which we can build:- a taxonomy of the tools used by taxonomists.

As this categorisation improves it can be used to
  • inform our understanding of process,
  • as a repository for collating tools and
  • a schema which we can analyse to understand the tasks, time and priorities.
Initially, we are collecting, analysing and seeking further information on the tools and the troubles or pleasures they bring.

Approaching from the Side (Breadth of consultation)

Practice and needs vary with discipline, taxa and institutions. As we improve our understanding, we hope these differences may become amenable to analysis. We plan to align the common elements of flowcharts (for the same process) tailored to different applications ( e.g. the collection and accession of lichens vs frogs). Similarly, we will collate and align what we can of terminology, technology and procedures in different disciplines and institutions.

As our understanding of differences improves, it may become possible to
  • transfer technologies
  • address common problems.
Similarly, process or procedural improvement may be transferable.

Workshops, Consultations and Interviews.

We are consulting as broadly as possible. As we are seeking information, we are not taking steps to achieve an even, unbiased sampling of taxonomic practice in Australia. Indeed, this would be counter-productive as the distribution of taxonomic effort is not uniform across the disciplines and is in some parts rather thin. Consultations take the form of structured workshops, repeated conversations on a theme and interviews. We will test the validity of our conclusions through iteration (i.e. further consultation). If it is merited, we may also formally test them through structured surveys.

The following table will give you some indication where information has been gathered.

(Needs updating)Nov 08

Taxonomic Process Discipline/Taxa Institutions Progress
Planning Botany, Vertebrates, insects ANWC, CPBR, AMI begun
Maintenance Data, collections   pending
Collecting Botany, Vertebrates, insects ANWC, CPBR, AMI underway
Processing     pending
Curation of ....      
Specimens Botany, Vertebrates, insects ANWC, CPBR, AMI underway
Research Products Botany, Vertebrates, insects, Zool ANWC, CPBR, AMI, Kango, UnimelbZ underway
Data Botany, Vertebrates, insects, invert. ANWC, CPBR, AMI,MOV pending
Collections Botany, Vertebrates, insects ANWC, CPBR, AMI underway
Research Botany, Vertebrates, insects ANWC, CPBR, AMI, UnimelbZ underway
Publishing Botany, Vertebrates, insects ANWC, CPBR, AMI / Flora of Australia begun

Analysis

Taxonomy (in its broadest sense) cannot be captured by any sort of Linear Flow charts (as detailed above). In its current state, our understanding also lacks the requisite definition and clarity (with respect to concepts, terminology, processes etc etc ) to be amenable to standard process modelling techniques; for that to occur further work is required. At the level of individuals taxonomists, many work practices (e.g alpha taxonomy or in the lab) are often linear & sequential. Certainly this is the case when applying specific techniques and even more so for Standard Operating Procedures.

To that end, this analysis is currently a four layer approach (to be revised as our depth of understanding improves). The most abstract layers allow us to capture important inter-relationships in the stages which make up the practice of taxonomy (and consequently common bottlenecks). In order from most abstract to least abstract the layers of analysis are:-

  1. Process Flow Model (showing relationships between Stages of the Taxonomic Process at a high level)
  2. (Near Linear) Flowcharts (for discrete and tractable elements of the Taxonomic Process)
  3. Methods / Techniques
  4. Standard Operating Procedures

Process Flow Models

At the higher conceptual levels the meaning of elements has been kept vague and ill-defined on purpose. In part, this is a reflection of the variability in the meanings of several terms in common usage. It is also as an initial approximation; allowing the freedom to fill in the details as they come to light.

There is also considerable overlap in what we are treating as discrete stages of the taxonomic process. As noted above, research in taxonomy is recursive. Research, Collection, and Curation all have key core activities in common. Indeed, key factors which drive activity in each of these endeavours are found in each other:-

e.g. curating a collection ---> a need for further collection of materials --> further research

research----> collection of materials --> further curation

episodic collection of materials --> further curation ---> further research

The representation of data, information and knowledge in these diagrams is limited to points at which data, information and knowledge accrues or is accesssed. This is because data, information and knowledge are the "products" of these processes. As such, details of the flow of information and data lie in a plane tangential to the taxonomic process.

Process_versus_Dataflow_3d_v1.gif

Interpretation of Process Flow Model Diagrams

Here is a key to the high level flow diagrams. The meaning of elements has been kept vague and ill-defined on purpose. In future iterations we hope to fill in the details and eventually to have a formal description of the taxonomic process.

key_2.gif

green boxes?

Boxes indicate what is hard to describe at this level. In the taxonomic systems diagrams, you will notice some parts of the process represented by boxes - indicating an item which is currently problematic or not a process. Of course, at some points TRIN's understanding of processes in taxonomy will be incomplete. We wish to identify where this is the case and to seek deeper understanding. In the flow diagrams, you will notice (mostly green) boxes - representing
  • information or data (& so a product of process not a part of it)
  • part of the process that is difficult to represent at this level
  • part of the process we are yet to understand (or even seems to be a black box process black box process). We will seek to minimise the number of these boxes which are not to do with information or data.

Flow Charts

The nearly linear parts of the taxonomic process (represented by flowcharts) are almost certainly amenable to standard process modelling techniques such as Business Process Models (if taken together with methods and standard operating proceedures).

We will assess whether this is worthwhile as the project proceeds.

FlowChart_legend.gif

Methods/ Techniques

Definition: A generalizable description of some discrete part of the taxonomic process.

A good exemplar being the Methods section of a scientific paper i.e. described in enough detail to allow replication under reasonable conditions.

It is not our intention to write taxonomic methods. Rather we will analyse extant and forthcoming methods. Many opportunities removing impediments to taxonomic practice lie here. Many of the Methods used in taxonomy share the same set impediments.

Examples..... to come

Standard Operating Procedures (SOP)

Definition: A description containing sufficient detail to replicate a task.

In the lab and field, taxonomy includes tasks with substantial logistical elements e.g management of expeditions or research products (e.g. DNA libraries). In the context of this project, this terminology is useful in that it allows us to address organisational issues which impede taxonomy.

SOP's are beloved by those who implement Quality control Systems and Safety Management Regimes. The term carries some negative baggage as a result. It is not our intention to write Standard Operating Procedures nor to force them upon others. We analysing explicit and implicit standard procedures to seek means to improve practice.

Project Status . %REDDOT{"ProjectOverviewProjectStatus"}%

3508761485 78393d7114.jpg
Workflow vs Dataflow

Review June 09

Progress on this analysis of Taxonomic Processes in Australia was put on hold from December 08 until June 09.

In the interim, there have been some developments of relevance in the changing mission of the ALA and our increasing understanding of what is required to manage molecular data for taxonomy. It seems appropriate to recast the analysis of the taxonomic process as more of a whole of system analysis.

Taken together
  • Margaret's Framework of Australian Taxonomic Datasystems
  • This analysis of Australian Taxanomic Processes
  • Taxonomic LIMS such as micro (molecular) LIMS?
    • Entity Relationship Diagrams
    • UML (may be overanalysis) - see below.
  • SOP's

orthogonal.jpg
Orthogonal

Are all elements of a greater whole viewed at different scales - but more than this they are aspects operating along different nearly orthogonal axes i.e.
  • Process (collection/research/curation)
  • Data (production/consumption/dissemination),
  • knowledge (creation/review/dissemination/transformation into data).

Strata comprising Taxonomic Processes (Progress)

Further attention is also due to the strata present in taxaonomic processes in Australia; especially along lines drawn by discipline, taxon, institution and lab. Within Discipline,

Progress by Stratum
Progress in Generalised models of process.

Taxonomic Process Progress Status indicative Comments
Planning begun ++ stub; linear flowchart; pinchpoints.
Collecting underway ++++ easily imagined; variable
Processing pending + stub
Curation of ....Specimens underway ++++ developed; further consultrations requied.
Curation of .... Research Products underway ++ Stub / see also & also?
Curation of ....Data pending ++ stub; further consultations req. writeup BL interview
Curation of ....Collections underway ++ stub; some consultations to document. Should not be hard to gather data
Research underway +++++ well underway. some challenges
Publishing .. scientific papers begun +++ lots or thougth required
Publishing .. to data services begun ++ links to EMC's framework; ALA; Name services; critical; requires care
Maintenance pending ++ stub

Need to Standardize analysis of each stage

The development of the project has been iterative. Content and processes reflect the bootstrap driven manner in which it was developed. Format and content was progressively refined as consultations progressed.Taken as a whole a more standard approach can now be applied.

Based on work to date the elements would be
  • A general flow diagram
    • carefully kept ambiguous and ill defined so as to be an overview uncomplicated by domain-details as possible. (actually I am unsure how successful I have been in keeping these diagrams uncomplicated).
    • Primary goal being to show the inter-relationships between discrete stages without getting hung up on the nature of these relationship (I think this has been more successful).
  • Background
    • Overview, summary of issues and themes
  • Steps
    • in so far as possible, a further breakdown of steps identified in the flow diagrams. Emphisis is on
  • Pinch Points.
    • Identified Solutions to Pinch points
  • Record of Consultations
    • need to include a table of who has been consulted and records of the outcomes from the consultations.

research methods & verification

Data collection

Qualititative Data

Almost all content gained to date has come from consultation both in workshops and from face to face meetings with specialists. This has proved a fertile source of data especially in discovering details which we need to document. Indeed essential.

A lot more qualitative data is required to flesh out the details which are yet to be documented especially for
  • Maintenance
  • Publishing
  • Curation
  • Processing
Work is still required across all stages to collect qualitative data. There are a lot of details which are still not present or needing confirmation with practictioners. That said, the processes in outline are generally simple and easily understood from a meeting. The specifics are what varies.
How will we know that we have a good representation of the taxanomic process?
Criteria.
  • i) representative consultations across discipline, taxa and institution.
    ii) convergence. No new big picture details
how much qualitative information is enough?

Quantitative

Time and motion analysis methods can be applied to ANIC and ANH .

Within Disciplines

analysis of specific in-discipline methods.
  • AMIT.
  • Cate Le Mann Anic
  • Beth

Within Taxa

Within labs

Molecular Lims could be extended to the lab practices of alpha taxonomy especially sorting.

Across Disciplines (Distinctions and differences)
This Taxonomic Process can be

UML

Components of a fully fleshed UML. The terminology is not a great match but the underlying concepts are useful. Tho' rigorous it can be put so much attention to detail as to be obsessive. All in all, still worth considering.

  • UML (may be overanalysis) quick overview
      • Structure diagrams
        • Class diagram
        • Component diagram
        • Composite structure diagram
        • Deployment diagram
        • Object diagram
        • Package diagram
      • Behaviour Analysis
        • Activity diagram
        • State diagrams
        • Use cases
      • Interaction Analysis
        • Communication Diagram
        • Interaction (overview) diagram
        • sequence diagram
        • timing diagram
Warning: Can't find topic TEndOfTopic