1. General
Category
Other
Please describe Other
Our solution address different SGD goals, such as Migration, Vaccine, Gender Equality, Vaccine and Good Health, to name a few.
Category
Other
Please describe Other
Our solution is meant for decision-makers from the public and private sectors (governments, multilateral organizations, and private companies)
2. Project Details
Company or Institution
IDB (Inter-American Development Bank)
Project
CivicLytics
General description of the AI solution
This AI tool allows you to understand complex civic issues by listening to the perceptions and concerns of millions of citizens in Latin America and the Caribbean in real time. We process and analyze open civic data that citizens express on the Internet (social networks, blogs, chatbox). From the IDB Group, we add this information by supporting governments, citizens, and the private sector to drive inclusive solutions to local and regional recovery.
Website
https://bidciviclytics.citibeats.com/en/
Organisation
Citibeats
3. Aspects
Excellence and Scientific Quality: Please detail the improvements made by the nominee or the nominees’ team or yourself if your applying for the award, and why they have been a success.
Citibeats’ algorithm analyzes large amounts of unstructured text data and other types of files where text can be extracted to identify in real time, social trends, people’s opinions, and concerns. All this information represents precious actionable insights for governments, multilateral organizations, and private companies to act upon and improve citizen’s lives.
Through the collection and processing of all this data, it is possible to obtain different outputs such as:
1. REAL-TIME ALERTS: data anomalies, relevant clusters, and on-demand cluster monitoring.
2. LEADING INDICATORS: social signals which tell you what will change, before you see it in traditional survey methods and the media.
3. SOCIAL INDEXES: social indexes providing stable information such as the level of trust in government from citizens.
Citibeats’ platform, thanks to its algorithms combining NPL (Natural Language Processing) and ML (Machine Learning), can filter relevant content, classify user opinions and information into categories, and extract actionable insights and patterns automatically; in real time; in changing environments; at scale; and reducing bias associated with gender.
Being language agnostic and culturally adaptable, our AI platform allows easily setting up projects across geographies by automatically integrating slang, dialects, and local expressions.
This versatility makes our platform the perfect fit for projects involving geographical areas or ethnic minorities that in normal circumstances would remain “invisible” and not have any influence on decision-makers.
The algorithms are designed to detect not just single keywords but to focus on complex texts. In particular, they are able to detect and distinguish any types of text differentiating among intents (narratives expressing complaints and questions) and sentiments (positive, neutral, negative).
Scaling of impact to SDGs: Please detail how many citizens/communities and/or researchers/businesses this has had or can have a positive impact on, including particular groups where applicable and to what extent.
Starting from March 2020, 21,775,202 comments in 5 different languages were analyzed. The analysis was run in 26 different countries tracking different COVID-19 categories to collect actionable insights within the SDGs framework.
In the Bahamas, Barbados, Guyana, Jamaica, Surinam, and Trinidad & Tobago the three most significant topics were:
1. After the reduction in the number of infections, concerns diversify
2. From supporting local businesses to supporting black-owned businesses
3. The Black Lives Matter movement focused criticism on the inequalities in the police force
In Argentina, Brazil, Chile, Paraguay, and Uruguay, the three most significant issues were:
1. Confinement triggers mental health problems
2. Neighboring countries with different economic concerns
3. Change of mentality towards individual responsibility of Brazilians
In Bolivia, Colombia, Ecuador, Peru, and Venezuela the three most significant themes were:
1. The flexibility in Ecuador generated peaks of doubt
2. Fear of contagion in urban markets in Peru
3. The Andean economies reinvented themselves after 3 weeks at home
In Belize, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Mexico, Panama, the Dominican Republic and Haiti, the three most significant topics were:
1. The closing of borders between Nicaragua-Costa Rica causes havoc
2. Gender gap marked by social isolation
3. Cash transfers could prevent famine
Scaling of AI solution: Please detail what proof of concept or implementations can you show now in terms of its efficacy and how the solution can be scaled to provide a global impact ad how realistic that scaling is.
Citibeats’ platform is a GDPR compliant SaaS license guaranteeing easy replicability, high scalability, and the capacity of adapting to the needs of the customer in a very responsive and adaptive way.
To that end, we have recently launched four package versions of our solutions, focused on Migration, Vaccine, Gender Equality, and Sustainability.
Citibeats’ platform has an intuitive interface and high degree of adaptability and does not require coding or data science so that anyone in your team can easily access and use it. This is just one of the characteristics that per se makes Citibeats’ platform easily scalable and “open” to any user.
Once a customer implements Citibeats’ solution, then it is very easy to run the same project across geographies or doing new research on a different topic, or analyzing/adding new categories.
The system, as already explained, learns fast. All the customer needs to do is specify the terms and limits of the project and the system will automatically give results. This explains why the model creation process via our platform is easily replicable and does not require starting from scratch every time that the customer wants to create a new project.
Ethical aspect: Please detail the way the solution addresses any of the main ethical aspects, including trustworthiness, bias, gender issues, etc.
The CivicLytics Citizen Observatory, through the use of AI to analyze Social Big Data, seeks to better understand the needs of citizens in real time and to reflect the dignity, trust, security, and visibility of people.
Citibeats (one of the first four tech startups to successfully implement the fAIr LAC tool) guarantees:
– Respect of the privacy
– Bias reduction (referred to the data itself, the training process of the AI, the extraction of insights, and the interpretation of those insights)
– Use Cases with a positive social impact (migration, health, natural disaster prevention, violence on women, and more)
Citibeats’ technology allows you to integrate different text data sources (Twitter, Facebook Public Pages, Instagram, Forums, Blogs, Videos, and News Comments) including data on demand for specific groups. Data on demand uses open distributed collaborations (or crowdsourcing) which provides visibility, knowledge, and information of more local situations. Crowdsourcing consists of a questionnaire (designed by Citibeats) sent to a defined network of verified citizens, who receive financial compensation for providing valid and complete answers to a full questionnaire using an application on their mobile phone. Our algorithms are designed to offer a representation of the whole population, reducing gender-related bias (due for example to the design itself of the algorithm, to its training and/or to the data itself) and calibrate the results of the research. In this way, diversity and inclusion are two main pillars of the models created through the platform.