Ever-changing Flags

The national flag is one of the symbols that help the formation and maintenance of the identity of a nation. Changes that occur in a country throughout history are often reflected on the design of its flag, whose elements bear meaning and are part of the country’s culture. Modifications to the design of a flag are normally sporadic and, in most cases, a flag remains unchanged for long periods of time. However, our society has now easy access to global information, which results in a sense of constant change. As such, in addition to the sense of identity, a nation can also be assigned a “mood” – i.e. what is happening in the country at the moment – and we believe flags can be also be used to represent it.

We present a system that generates flags based on trending topics of countries, retrieved from real-time news. These topics are used to drive a process of visual blending that alters the original flag of the country. In this sense, the produced flags can be seen as visual representations of the current “mood” of the country.

**Figure 1**
Examples of flags generated on November 15th 2019. Below each flag, the country of the original flag and the trending topic used in the generation are identified.

Data

There are three kinds of flag data: visual, e.g. a dataset of fully scalable vector graphics of flags (flag-icon-css); semantic, i.e. meaning of flags; and about flag structure, e.g. the platform Flag Identifier. As our goal is to produce flags by transforming existing ones, the most useful types of data are semantic and visual. Since we could not find any dataset that associated both types of data, we decided to produce one.

Generating Flags

One of our main goals was to be able to maintain the resemblance with the base flag, allowing the identification of the country. For this reason, our system was grounded on two base assumptions: for each flag production, an existing flag would be given as input and the transformations should not go beyond the point in which the original flag is not recognisable anymore – i.e. the produced flags should not be seen as a totally new flag but as a transformation of the original one. At a first stage, the process of producing flags involves the search for elements that match a queried word, which are then used to transform the original flag. The search is conducted in three different places:

Existing Flags: as already described, we produced a flag dataset that included both visual data and semantic data. Using this dataset, a search for the input word is conducted on the meanings associated to elements of existing flags;
Colour Names: we merged existing datasets to produce a list of 3,476 colours and associated names (e.g. #ef4026 has the name “Tomato”). This list is used to search for the input word;
Emoji: we use the EmojiNet [1] dataset to find emoji based on the input word (similar to what is done in Emojinating system).

The transformation made to the flag depends on the type of elements found. For example, if the input word is found on a colour name, the colour is applied to an element of the base flag. On the other hand, if the element found is an emoji, it can either be added to the flag or replace an existing symbol.

Trend-driven Flags

The notion of “mutable flag” gains even more significance when combined with a sense of reactivity – something is “reactive” when it changes according to external input. Our main goal was to generate flags that changed according to current events, instilling a quality of “being alive” into them. As such, when generating a flag for a given country, the system automatically collects the latest news titles in English that mention the country’s name and identifies the most predominant topics (nouns). From these topics, we then generate flags that represent the “mood” of the country.

**Figure 3**
Flags and produced explanations, automatically generated on 15/11/2019

Producing Explanations

Flags have a huge layer of symbolism as most of their elements have associated meanings. When generating a flag, such layer is also of great importance. As such, in addition to producing flags, we also provide explanations that provide clues of how and why the flag was changed. This creates a connection between shape, meaning and explanation, which, we believe, serves to provide a strong conceptual ground for the produced flag. Explanations follow a predefined structure:

[element X] represents/stands for/symbolises [Y]

where Y is the queried word and X depends on the change nature.

Perception by users

Even though flags are publicly displayed, their meanings are often only understood by people that are familiar with them. By changing an element in a flag we also change the associated meanings – e.g. a colour replacement carries a meaning, which will be assigned to the flag. In [2], we conducted a user study with 16 participants to assess the perception of generated flags. The results show that the participants can identify the original flag, but they have certain difficulty in identifying the meaning of the changes applied to the flags. The user study also points to the potential of generated flags to raise awareness to a certain events.

Applications

The developed system is only a starting point for several applications. Our main goal is to develop artefacts that foster a discussion on what a nation’s identity can encompass and how the characteristics of current society can be exploited. At the moment, we have used the system in two different artefacts: a web-based application and a real-time installation.

Impact and Ethics of Modifying National Symbols

The limits of use of a national flag have long been a topic of debate. Flags are prone to be misappropriated and are often used in acts motivated by political reasons – e.g. flags being burnt in protests. Several cases exist of controversy around what is considered legal and what is to be seen as flag desecration. In [3], we question to what extent do flags actually represent constantly evolving nations when they are subject to rules often against change and transformation.

More information

References

[1] Wijeratne, S., and Balasuriya, L., Sheth, A., Doran, D. 2017. EmojiNet: An Open Service and API for Emoji Sense Discovery. In 11th International AAAI Conference on Web and Social Media (ICWSM 2017). Montreal, Canada; 2017.

Publications

J. M. Cunha, P. Martins, H. G. Oliveira, and P. Machado, “Ever-changing Flags: Trend-driven Symbols of Identity,” in 8th Conference on Computation, Communication, Aesthetics & X (xCoAx 2020), 2020.
- Bibtex
- |
- PDF
J. M. Cunha, P. Martins, and P. Machado, “Ever-changing Flags: Impact and Ethics of Modifying National Symbols,” in Proceedings of the Eleventh International Conference on Computational Creativity, ICCC 2020, Coimbra, Portugal, September 7-11, 2020, 2020, pp. 410-413.
- Bibtex
- |
- PDF

Author

João Miguel Cunha

Hugo Gonçalo Oliveira

Penousal Machado