community logos (iteration 1)

a project exploring ideas for community logos

Introduction

This page presents a snapshot of some of the work undertaken in relation to a current personal project concerned with exploring ideas for potential logos for use by the nascent Neuro-Symbolic Systems AI research community that is emerging within the Computer Science department at City, University of London.

Logo theme A

This section lists a few versions of a particular logo design idea that I refer to as theme A.

Some considerations motivating the design idea expressed here:

  • neuro-symbolic systems is a general concept that covers many different approaches to neuro-symbolic AI; this observation inspired an umbrella metaphor
  • the neural space is well-characterised by deep learning alone, hence it is represented by a single icon
  • the symbolic space takes many forms, each having a distinct character; hence the symbolic space is represented by several different icons
  • the deep learning icon combined with (one or more) symbolic icons is meant to represent a particular family/approach/discipline of neuro-symbolic AI and the research interests/specialisations of the academics that associate themselves with that mode of neuro-symbolic AI; for example:
    • DL + Logic ==> LTN, Logic Neural Networks (IBM), …
    • DL + ILP ==> CILP, …
    • DL + natural language ==> NLP
    • DL + KGs/SW ==> some catchy moniker yet to be determined, …
  • the logo idea is designed to readily accommodate community growth whilst still permitting individual specialisations to be visually represented
    • new symbolic space icons can be introduced under the umbrella canopy, over time, as the community grows to embrace new, emerging disciplines/specialisations, and without requiring successive, major logo redesign/restructuring efforts that would confuse the public and damage the brand
    • there’s already room for a 6th icon; what’s missing ?
    • growth would occur downward with the introduction of successive rows of symbolic icons: row 3, row 4, …

Thoughts on NLP:

  • NLP is a tricky discipline to represent as distinct neural and symbolic elements because its well-known moniker, ‘NLP’, already embraces deep learning. So, in the logo, the symbolic icon for NLP emphasizes its symbolic component, natural language, and uses Latin for this purpose in an attempt to suggest language-agnosticity. For now, the symbolic icon still refers explicitly to ‘NLP’ because this moniker is so well-recognised.
  • However, the option exists to replace the icon name ‘NLP’ with ‘natural language’ and to let the discipline of NLP be represented implicitly via the combination of its ‘deep learning’ (neural) and ‘natural language’ (symbolic) components.

Viewing the logos

  • To see the logo images better, increase their size by clicking the + icon in the menu bar at the top of their enclosing frames. 250% works well for me on my machine.

A-7

Here we're trying to use colour to distinguish the neural and symbolic spaces: brown for neural, beige for symbolic. The brown-to-beige colour gradient within the umbrella canopy is meant to connote the merging/combination of the two spaces, or neuro-symbolic *integration*, or some such. The position of the umbrella shaft/handle is meant to be a visual, spatial delimiter separating the two spaces, with neural (brown) to the left and symbolic (beige) to the right. The shaft is short so as to serve this 'spatial delimiter' purpose whilst remaining fairly unobtrusive. Its shortness and off-centre positioning also serve to lend an unexpected 'quirky' character to the image that some may find attractive. The brown (wooden) handle is for fun; it just, well, kind of works.

A-8

Same as A-7 but with a modified community name/description. The community name now has a smarter, traditional font (Copperplate) and uses two colours rather than one. The name is underlined and accompanied by a suggested community slogan/tagline. But while the colour of the slogan/tagline matches the background, the result is too pale and weak.

A-9

Same as A-8 but now the community slogan/tagline has a bolder colour (same as the word SYSTEMS) and is tighter to the line under the community name.

A-10

Same as A-9 but now the brown-to-beige colour gradient filling the umbrella canopy has been replaced by the background colour. This permits an experiment with a 3-colour community name where we use the colours of the words NEURAL (brown) and SYMBOLIC (beige) to echo and reinforce the distinction between (and combination of) the neural and symbolic spaces. But the beige lettering of SYMBOLIC is a bit too pale, at least against the current background colour. This 3-colour community name idea could be more effective if explored in the context of a different background colour or in the context of different overall colour schemes.

Logo theme B

This section explores a variation of logo design theme A, called theme B.

The logo idea of theme A relies on the logo being displayed with a size large enough for the detail of the individual neural/symbolic icons to be visible/readable. This is unlikely to be feasible in all potential use cases. So theme B explores a variant of theme A whose design may be better suited to space-constrained use cases.

Allowing the community’s identity to be represented by both logo themes A and B could, potentially, work without damaging the community’s brand/identity. The umbrella icon in theme B clearly echos and refers to its older brother, theme A, and should be immediately recognisable and interpretable to those already familiar with logo theme A.

B-1

Version B-1 retains the 3-colour community name, in both the written name and the letters within the umbrella canopy. This serves to highlight the distinction between (and combination of) the neural and symbolic spaces. But, again, the beige colouring for 'S' and 'SYMBOLIC' is a bit too pale.

B-3

Same as B-1 but with a 2-colour community name, using black for both 'N-S' and 'NEURAL-SYMBOLIC' to help them stand out boldly.

B-4

Same as B-3 but reintroducing the brown-to-beige colour gradient within the umbrella canopy. This allows the colouring of the umbrella canopy to express the distinction between (and combination of) the neural and symbolic spaces, which relieves the community name from having to perform this task.