Related to participatory mapping, social mapping and VGI, we can find a diversity of theories related to citizen-led data gathering and visualization projects. Maps have been great tools, designed to visualize information with a high level of complexity in a plane. The metaphor of the millennia was that – with the internet – map possibilities will be overtaken with the application of different visualization systems and more complex data management tools.

But maps present some difficulties which are often forgotten. Complexity is only taken simple when a reader can interpret it. In other words, the reader has learned to interiorize map symbols and aesthetics codes. The spatial metaphor also requires higher skills, when it is about to cover big territories or topographies, or when a cartographic artifact should guide us in a spatial navigation.

Beyond human abilities, another issues to deal with have to do with the technological infrastructure. The speed required to download the tiles, devices required to visualize them, or the speed connection required to get real-time information, are not always feasible options. Here, we are talking about cases where connectivity is null or there isn’t access to devices, situations that could happen in refugee camps; or where a person should pay attention to a specific task, like driving a car; or when contextual information is unknown by the user and he/she should learn it very fast in order to understand what is going on. Even a more rare but possible scenario: a person who doesn’t know much about the place or the code which the map is showing information about.

After decades of trying to give artificial intelligence (AI) some kind of utility for people, conversational bots or chatbots, and virtual assistants have appeared . In my research, I’ve analysed all the bot systems currently available. During the last year and 2016, new researches related to how these systems can help to specific topics or human/social tasks were published (Luger & Sellen, 2016). Furthermore, platforms for creating bots have been proliferated giving an answer for new situations, including social networks like Facebook, Telegram or Slack. They provide an API to access to their message platform to program bots that respond to some user text patterns.

Despite is a promising path, the development of chatbots is clearly limited. Challenges are the correct development of Natural Language with local modifications of languages and expressions (1); the increase of Machine Learning processes to improve the learning in more intelligent way (2); and the development of new functionalities (3). Regarding the latter, the current exploited functionality is to develop simple tasks, like scheduling meetings or requesting an Uber car, or find relevant info for buying tickets.

In a little period of time, chatbots are becoming alternative interfaces to the “traditional” mobile applications. This is happening because of the simplicity of the information provided and the increasing capacity of learning systems to improve its efficiency. The answer provided from bots is highly contextual and easy to understand, and it opens a new prospect, not only for service companies like Uber or Amazon (as they can sell more, in an easy, fast and subtle way), but also for being used in Georeferencing and Participation systems.

I’m just thinking about some useful ideas. For example, governments can use bots to chat with their citizens without the necessity of having and operator behind. Survey systems or formality tracking systems, can be easily managed by bots without forcing people to queue at any public building. They can also provide user orientation virtually. For giving a real case scenario, Polis is developing a service for governments that use AI for developing surveys. (https://pol.is/gov)

On the other side, for georeferencing services, bots can provide contextual information after a user request without rendering a map. Moreover, they can return images (like street view), or graphical analysis, or image maps from a specific topic that the user asked for. Regarding social maps, using a provided API, bots could be generated to gather or publish data in real-time without the need of new mobile apps or complex interfaces. A particular example case is the Telegram bot for refugee camps’ volunteers in Spain. https://twitter.com/canalrefugiadxs/status/752523286165254144

It is clear that chatbots used in text platforms could be more uncomfortable to use than speech recognition virtual assistants like Siri or Alexa, but those systems are not very precise in their answers and speech understanding. For that reason – and because predictive keyboards are getting more effective and providing better feasibility (Kristensson, 2015) – bot platforms become an interesting tool to develop. In addition,  it is also possible to take advantage of the smartphone’s camera using image recognition or computer vision in combination with chatbots

Despite the idea that we are getting stuck in an IVR (Interactive Voice Response) system with asteroids as Bill Evans points out (http://ben-evans.com/benedictevans/2016/3/30/chat-bots-conversation-and-ai-as-an-interface), this is a step forward to develop more advanced systems in virtual assistants.

Finally, we have to remember that the adoption of a specific technology does not only rely onits functionality, but in its relation with society. We don’t have to dismiss all the technologies that are recurrently appearing one after another, because this is the way that they will get stabilized (Bijker et. al., 2012). In the meantime, we should take the opportunity of using available tools to provide better services to people, expand the research of new social phenomena and take new challenges as well.

Bijker, W. E., Hughes, T. P., Pinch, T., & Douglas, D. G. (2012). The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. MIT Press. Retrieved from https://books.google.es/books?id=B_Tas3u48f8C

Kristensson, P. O. (2015). Next-Generation Text Entry. Computer, 48(7), 84–87. http://doi.org/10.1109/MC.2015.185

Luger, E., & Sellen, A. (2016). “Like Having a Really Bad PA” In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems – CHI ’16 (pp. 5286–5297). New York, New York, USA: ACM Press. http://doi.org/10.1145/2858036.2858288

Manuel Portela, GEO-C PhD student