User Scenarios for Client and Mapper Design

From Filtered Push Wiki
Jump to: navigation, search

Definitions

FP Network Software
The FilteredPush network software. Allows the deployment of FilteredPush network instances.
FP Network Instance
An instance of a network formed by cooperating nodes using the FilteredPush Network Software along with clients to that network.
Node
A deployment of the FilteredPush Network software at a site configured to participate in atleast one FilteredPush Network Instance. The node provides at least an interface allowing clients to interact with the FP Network Instance and a backing component (triage in the prototype) to allow the node to utilize capabilities in the network to react to messages from clients. A node may include other network capabilities.
Capability
Network knowledge, analysis, or message handling functions present at some but not necessarily all nodes.
Client
A piece of software that interacts with a FilteredPush Network Instance through a defined programming interface. Clients may or may not also be Data Providers.
Data Provider
A client to a FilteredPush Network Instance which assumes responsibility for the data curation of an authoritative copy of one or more data sets that are made available to the network instance. A Data Provider registers with and is available to an instance of a Filtered Push Network.
Applecore
Guidance for the use of the TDWG DarwinCore within the botanical community.
A2C-net
An instance of a FilteredPush Network deployed in the botanical community using the Applecore guidance and the TDWG DarwinCore as a standard vocabulary. In A2C-net, unredacted, unmodified locality and georeference data is provided to the A2C-net network instance by Data Provider clients, with that data exposed only to other authorized clients.
A2C-net system
The A2C-net, it's registered data providers, and it's registered clients.
Specify
Specify 6 as produced by KU along with FilteredPush plugin components produced by the FilteredPush project.
Specify Workbench
Specify 6 Workbench as produce by KU along with FilteredPush plugin components produced by the FilteredPush project.
Duplicate Set
A set of records of botanical duplicates along with the annotations that have been applied to this set and a synthetic consensus record produced by the A2C-system.
Consensus Record
The synthetic consensus record that is part of a Duplicate set which represents the best current understanding of the A2C-system of a TDWG DarwinCore per Applecore record that represents the information that is shared in common amongst this duplicate set (Locality, Georeference, Collector, CollectorNumber, Collector's original Determination, Current Determination, etc.) but not institution or specimen specific information (Accession Numbers, Barcode Numbers, determination histories, etc).

Scenarios


(1) Find Duplicates Specify Demonstration for SPNHC 2011


(A) I'm working on X

A person starts Specify, and through a user interface, indicates that they will be working on a particular taxonomic group or geographic region [X].

Specify launches a message into A2C-net, through the API exposed at a node, telling the network that a user at this Specify instance will be working on X and that they will expect rapid response times on find duplicate queries about X.

(B) Find Duplicates (Native Specify)

A person starts Specify, creates a new CollectionObject record, and in the Collection Object form, enters a collector name, collector number, and date collected from a herbarium sheet or herbarium sheet image. [Alternatively, the entry of a new collection object starts with a modal dialog that requests these three specific fields, and either finds a duplicate or populates these fields on the collection object form.]

If a potential duplicate for these terms is known to the A2C-system, the system responds in less than 0.1 seconds from the date collected field loosing focus by bringing up a dialog that displays the consensus record(s) that may be potential matches to this specimen, and allows the user to examine the underlying data records and provenance for the consensus record(s).

If no potential duplicate is found, the user keeps transcribing label data from the sheet or image into the fields on the Collection Object form.

This user interface behavior is customizable, such that a specify user can indicate which concepts are involved in what duplicate findings query (e.g. specimen finding, locality finding), and when the find duplicates query is triggered (e.g. button press, each time any of a set of fields lose focus).

If a user decides that a potential duplicate matches the data they have on the sheet in front of them, they can accept the consensus record and have it automatically populate the fields on the collection object data entry form (collecting event data, locality data, earliest determination, preparation type, etc.) The user can then fill in collection specific information such as barcode number. The user can alter information that has been passed from the consensus record to fields on the collection object form. When the user saves the botanical duplicate collection object record, one or more annotations are generated to the filtered push network, one adding this record to the duplicate set, others potentially annotating the consensus record with differences between this record and the consensus.

In the demonstration scenario, a particular field (e.g. county name) contains a value which is inconsistent in the consensus record and on the sheet in front of the user. This inconsistency results in the A2C-system generating an annotation (Annotation 1) on the relevant duplicate set pointing out the inconsistency and suggesting a corrected value based on the specimen at this institution.


(C) Find Duplicates (Specify Workbench)

A person starts the Specify Workbench, opens a dataset, creates a new workbench record, and in the workbench form or grid, enters a collector name, collector number, and date collected from a herbarium sheet or herbarium sheet image.

If a potential duplicate for these terms is known to the A2C-system, the system responds in less than 0.1 seconds from the date collected field loosing focus by bringing up a dialog that displays the consensus record(s) that may be potential matches to this specimen, and allows the user to examine the underlying data records and provenance for the consensus record(s).

If no potential duplicate is found, the user keeps transcribing label data from the sheet or image into the fields on the workbench form or grid.

This user interface behavior is customizable, such that a specify user can indicate which concepts are involved in what duplicate finding query (e.g. specimen finding, locality finding), and when the find duplicates query is triggered (e.g. button press, each time any of a set of fields lose focus).

If a user decides that a potential duplicate matches the data they have on the sheet in front of them, they can accept the consensus record and have it automatically populate the fields on the worksheet form or grid (collecting event data, locality data, earliest determination, preparation type, etc.) The user can then fill in collection specific information such as barcode number. The user can alter information that has been passed from the consensus record to fields on the worksheet.

When the user (saves? ingests?) the botanical duplicate collection object record, one or more annotations are generated to the filtered push network, one adding this record to the duplicate set, others potentially annotating the consensus record with differences between this record and the consensus.

(3) Ingest New Determination Annotation scenario.


(A) Inject Annotation

A researcher views an image of a specimen in Morphbank. The researcher asserts that the specimen is incorrectly identified, and with characters visible in the image as evidence, asserts a new determination for the specimen. Morphbank injects this new determination as an annotation into the A2C-net (Annotation 2).

The annotation contains: Darwin core terms for scientific name, scientific name authorship, the determiner, the date determined. A motivation section containing a description of the morphological evidence. An expectation Expectation_Insert. The user who created the annotation and the date the annotation was created. Annotation is similar to http://www.etaxonomy.org/ontologies/ao/aod_example_6.owl or http://www.etaxonomy.org/ontologies/ao/aod_example_4.owl

(B) QC Name

The A2C-system examines the taxon name asserted in Annotation 2, compares it against taxonomic authorities, and annotates the annotation with an annotation asserting that the name is found in IPNI, and the IPNI LSID for the name is provided.

(C) Ingest annotation

A data curator at the institution holding the specimen re-determined in (3.A) opens a user interface to view incoming annotations relevant to their collection. Among the annotations of most interest presented to them is the annotation produced in (3.B) (Annotation 2). This annotation asserts a new determination for a particular specimen record in their collection.

The data curator views this change in the context of the specimen record.

The data curator examines the provenance and evidence of the annotation, and accepts the annotation.

The change proposed in the annotation is made by adding a record to the determination history of the relevant specimen (with a new taxon being added if needed).

The data curator's action results in an annotation indicating their acceptance of the annotation being injected into the network.

Also available to the data curator is the ability to reject this annotation, or establish a filter based on this annotation where by future matching annotations will be accepted, or rejected, or ignored.

(D) Filter annotation on duplicate

A data curator at an institution holding a member of the duplicate set of the the specimen re-determined in (3.A) opens a user interface to view incoming annotations relevant to their collection.

Among the annotations of most interest presented to them is the annotation produced in (3.B) (Annotation 2). This annotation asserts a new determination for a duplicate particular specimen record in their collection.

The data curator establishes a filter whereby the A2C-system will automatically ignore any future annotations sent to that institution which contain a new determination on a member of a duplicate set where the annotated specimen is not held by that institution (but will still present other annotations on duplicates to the data curator).

Not specified, but needed

(A) A data manager maps a local collections database using a data mapping tool and joins the A2C-System.

(B) A user of the A2C-system views their interests and filters on those interests and can add new interests, remove interests, modify interests, add filters, remove filters, and modify filters.

(C) Add a node to the A2C-net.

(D) Failover mechanisms when a node in a filteredpush network is unavailable.