Appeal to scientists from the community antclub.ru

Community and Forum → Taxonomy. Classification → Appeal to scientists from the community antclub.ru

antfarmer, 10.02.2011 10:28

Hello!

I am the site administrator http://antclub.ru.

I have long been interested in ants, although at the level of an amateur, but I have reread all the available literature.

I want to make a convenient ant detector, but I don't have enough knowledge and experience in this area.

I am a programmer, and while preparing literature for publication on the site, I noticed that all the tables/determinants of ants are basically written in "one language" (i.e. using the same phrases/concepts) within the same book/author. But different authors often use different words to describe the same thing. Plus, it's not entirely clear what is more "slightly", "slightly" or "slightly".

It seems to me that all these descriptions can be easily formalized/unified - turned into a set of links to pre-prepared reference books of concepts.

For example, the reference book " A " - contains the terms of the subject of the description:

A - subject of description
- - - - A1-egg
- - - - A2-larva
- - - - - - - A21 - 1st age
.....
--------A25-5th age
- - - - A3-adult
----A31-male
--- - A32-queen
- - - - A33-worker
--- - - - - A331-soldier
- - - - - - - A332
-gamergat--- - - - - A333-pleergat

a reference book " B " - describes the structure of the ant:

B-ant
- - - - B1-head

--------B11-usyk
------------B111
-scapus----------------B1111-base
------------B112-flagellum
-------- - - - - - - - - B1121-segment
------------B113-mace

-------- B12-compound eye
------------B121-facet

--------B13-simple eye
--- - - - - B14-nape
- - - - - - - B14-platypus
... ... ... ...
- - - - B2-chest
- - - - B3-stalk
- - - - B4-abdomen
- - - - B5-legs
-- - - B6-wings

(I will not give the full structure, the idea is already clear)

In the same way, reference books "color", "pubescence", "signs" (size, shape, distortion, position, quantity, type of surface), "measurements and indexes", etc. are prepared.

As a result, the description of any ant can be reduced to a certain formula consisting of "phrases": "species":"from whom":"what":"which one" [:"by how much"].

for example:
"The head of a worker is completely sculpted, with wrinkles and shagreen." (Leptothorax alinae)

"Head" - WHAT (from the reference book "structure")
"worker" - From WHOM (from the reference book "subject of description")
"completely" - omitted (since this is already implied) or converted to the coefficient "BY HOW MUCH" (the strength of the sign's manifestation)
"sculpted" + "wrinkles" + "shagreen" - WHICH ONE (from the reference book "signs: surface type"), for example, we assign them the codes E11, E12,E13, respectively

and all this turns into the formula: Leptothorax alinae: A33:Б1:(E11,E112,E113):"completely" (in the car, of course, it is more convenient to store three entries, without brackets).

This will provide a few clear utilities:

- the most important thing is to keep all descriptions in a uniform form, this will eliminate the influence of the description's features on the description, discrepancies,linguistic features, etc.

- ability to compare descriptions of ants, analyze the structure of descriptions,
find common features, etc.

- automatic generation of tables/qualifiers (including in any language - you just need to translate reference books)

- ability to link schematic images to each reference point and automatically generate a schematic image of a specific ant, or a diagram of differences between several descriptions

- the ability to use the diagram to find suitable descriptions of ants, or just graphic schemes-determinants (for example, as in" Ants of the Moscow region " by Evgeny Anatolyevich Dunaev)

- by adding coordinates to the description, you can track geographical variability

This principle can be extended not only to ants.

My modest knowledge is not enough to implement this idea. You need a scientist who understands the subject well, has experience in drawing up descriptions and organizing them. The more correctly and accurately (and conveniently) the reference books/classifiers are designed and prepared, the more versatile and useful the tool will be.

To implement this idea, you need a SUPERVISOR (curator, consultant).

For my part, I am ready to devote all my free time to this project as a programmer, designer, coordinator and anyone else you need. Ready for serious work. I am ready to attract additional resources: I am sure that many ant lovers will be very happy to participate in something real and will take the assigned work with responsibility.

I will be glad to hear any comments/suggestions/criticism and any expressions of interest in this project in general.

Sorry if I wasted your time.

Sincerely, antfarmer.

Comments

19.02.2011 18:08, PVOzerski

Uw. antfarmer, I will try to say something-although I myself am not a myrmecologist at all (but there are general considerations).

If you take a close look at the definition tables, you will find that the range of features involved in determining them is not only wide, but also not entirely predictable. In other words, we add another type to the database - and it may happen that the attributes necessary to distinguish it from others have not yet been used in the tables. And immediately you will need to review the entire structure of the "directory".

I have, however, such an assumption that, in principle, your whole idea can be implemented in some DBMS via SQL queries. You just need to think through the structure of the tables.

Likes: 1

23.02.2011 13:01, antfarmer

Hello!

Thank you for responding!

2 PVOzerski:

"If you look carefully at the definition tables, you will find that the range of features involved in determining them is not only wide, but also not entirely predictable." - I have very carefully studied all the definitions/descriptions of ants available to me... and I think that the range of attracted features is very predictable!.. and it can easily be "squeezed" into hierarchical directories...

"And you will immediately need to review the entire structure of the reference book" - that's why I suggest separating different concepts into different reference books... even if an ant appears with some new body part, it, this body part, will grow on some other part of the body, and the reference book can be supplemented without changing the structure, and this will not affect the already encoded descriptions in any way...
even if an ant appears with another eye on its back , it will be a new eye that will be entered in the appropriate place in the corresponding directory...

this concerns possible structural changes... and all other descriptions of features can be reduced to distortions of previously known forms/parts... for example, an ant will appear with a new spike shape - this can always be described as a change in the shape of a standard spike... the number of changes in size/shape is quite enumerable, provided that the concepts of "what":"what"[:"by how much"]are separated

2 Proctos: in my opinion, Lucid is quite primitive - it uses only "flat" sets of primitive features (yes/no, number)... and although it is possible to group attributes into a hierarchy, this hierarchy does not carry any useful load...

...of course, you can also make a determinant in Lucid... but Lucid is a two-dimensional (feature x entity), and there is no way to combine features with each other, so you can only choose what is hard-coded in advance...

I propose to encode the features in several dimensions with an exact reference of each dimension to some type (body part, color, distortion, etc.)... this gives a huge number of feature variations, including for describing non-existent (still/already) ants...

2 PVOzerski: "I have, however, such an assumption that, in principle, your whole idea can be implemented in some DBMS via SQL queries. You just need to think through the structure of the tables." "I'll say it again... I am a programmer and I have no problems with either DBMS or SQL... A RESEARCH SUPERVISOR is required.

good luck.

This post was edited by antfarmer - 02/23/2011 13: 08

21.02.2012 17:37, antlook

As a rule, work on the taxonomy of any group of organisms and their classification is carried out in a certain sequence. This sequence involves ordering data arrays by creating card files. Card files are formed on various media, including paper ones (cards), as we can see for example in the library catalog, where the UDC (universal code) plays the role of a key.When choosing cards, we sort them by authors, sections, and year of publication. This is in the general case, in our particular case, file cabinets are usually formed that include the following general catalogs:
Diary (diary of entomological fees)
Author's catalog (list of references)
Systematic catalog (list of species)
Data processing (comments on sections or collecting facts, with links to sources)
To meet all the conditions, when creating an electronic application, you must meet the following conditions:
For the data format-
ability to use the accepted format in various applications
For the application-
High processing speed
Card filling
Ability to run from removable media (Portability)
The ability to work with large arrays
of most of the requirements is fully satisfied by TabletaPro http://zgsprojects.narod.ru/ where it is possible to create ordered lists of a certain structure, and its use along with other applications (Excel * - pivot tables, Access* - queries, GIS - visualization in space) significantly speeds up processing.
*these applications are a class of programs with a tabular data structure that can be implemented using free software (OpenOffice, LibreOffice, etc.)