The Armed Conflict Location & Event Data Project (ACLED) is a US-registered non-profit whose mission is to provide the highest quality real-time data on political violence and demonstrations globally. The information collected includes the type of event, its date, the location, the actors involved, a brief narrative summary, and any reported fatalities. ACLED users rely on our robust global dataset to support decision-making around policy and programming, accurately analyze political and country risk, support operational security planning, and improve supply chain management.ACLED’s transparent methodology, expert team composed of 250 individuals speaking more than 70 languages, real-time coding system, and weekly update schedule are unrivaled in the field of data collection on conflict and disorder. Global Coverage: We track political violence, demonstrations, and strategic developments around the world, covering more than 240 countries and territories.Published Weekly: Our data are collected in real time and published weekly. It is the only dataset of its kind to provide such a high update frequency, with peer datasets most often updating monthly or yearly.Historical Data: Our dataset contains at least two full years of data for all countries and territories, with more extensive coverage available for multiple regions.Experienced Researchers: Our data are coded by experienced researchers with local, country, and regional expertise and language skills.Thorough Data Collection and Sourcing: Pulling from traditional media, reports, local partner data, and verified new media, ACLED uses a tailor-made sourcing methodology for individual regions/countries.Extensive Review Process: Our data go through an exhaustive multi-stage quality assurance process to ensure their accuracy and reliability. This process includes both manual and automated error checking and contextual review.Clean, Standardized, and Validated: Our data can be easily connected with internal dashboards through our API or downloaded through the Data Export Tool on our website.Resources Available on ESRI’s Living AtlasACLED data are available through the Living Atlas for the most recent 12 month period. The data are mapped to the centroid of first administrative divisions (“admin1”) within countries (e.g., states, districts, provinces) and aggregated by month. Variables in the data include:The number of events per admin1-month, disaggregated by event type (protests, riots, battles, violence against civilians, explosions/remote violence, and strategic developments)A conservative estimate of reported fatalities per admin1-monthThe total number of distinct violent actors active in the corresponding admin1 for each monthThis Living Atlas item is a Web Map, which provides a pre-configured view of ACLED event data in a few layers:ACLED Event Counts layer: events per admin1-month, styled by predominant event type for each location.ACLED Violent Actors layer: the number of distinct violent actors per admin1-month.ACLED Fatality Estimates layer: the estimated number of fatalities from political violence per admin1-month.These layers are based on the ACLED Conflict and Demonstrations Event Data Feature Layer, which has the same data but only a basic default styling that is similar to the Event Counts layer. The Web Map layers are configured with a time-slider component to account for the multiple months of data per admin1 unit. These indicators are also available in the ACLED Conflict and Demonstrations Data Key Indicators Group Layer, which includes the same preconfigured layers but without the time-slider component or background layers.Resources Available on the ACLED WebsiteThe fully disaggregated dataset is available for download on ACLED's website including:Date (day, month, year)Actors, associated actors, and actor typesLocation information (ADMIN1, ADMIN2, ADMIN3, location and geo coordinates)A conservative fatality estimateDisorder type, event types, and sub-event typesTags further categorizing the data A notes column providing a narrative of the event For more information, please see the ACLED Codebook.To explore ACLED’s full dataset, please register on the ACLED Access Portal, following the instructions available in this Access Guide. Upon registration, you’ll receive access to ACLED data on a limited basis. Commercial users have access to 3 free data downloads company-wide with access to up to one year of historical data. Public sector users have access to 6 downloads of up to three years of historical data organization-wide. To explore options for extended access, please reach out to our Access Team (access@acleddata.com).With an ACLED license, users can also leverage ACLED’s interactive Global Dashboard and check in for weekly data updates and analysis tracking key political violence and protest trends around the world. ACLED also has several analytical tools available such as our Early Warning Dashboard, Conflict Alert System (CAST), and Conflict Index Dashboard.

Context

This dataset contain data about armed conflict locations & event data in Bangladesh from the beginning of 2001 until Novmber 2021. The 'fatalities' feature can be used as a target to model for predictions.

Acknowledgements

We thank ACLED for providing this data. Find ACLED here : https://acleddata.com/#/dashboard

Licence

Non-Commercial Licenses - ACLED’s full dataset is available for use free of charge by noncommercial entities and organizations (e.g., non-profit organizations, government agencies, academic institutions) using the data for non-commercial purposes, subject to these Terms of Use. Non-commercial licenses may also be granted to for-profit media outlets or journalists citing ACLED’s content in works of journalism; provided that such works are made available to the general public and benefit public discourse on the topic, subject to ACLED’s prior, written approval.

Inspiration

How many fatalities based on event type and subtype? What to expect when each actor is involved in conflict? What regions are impacted the most? What are the events that manifest more fatalities? Can we model and predict fatalities based on the features we have? Can we forecast the upcoming year's crime rate?

This dashboard is part of SDGs Today. Please see sdgstoday.org.Armed conflicts arise from many sources, including border disputes, civil war, and religious and tribal clashes. Increasingly, these conflicts are originating due to poor environmental conditions, such as lack of access to water resources and arable land, drought, and famine. The Armed Conflict Location & Event Data Project (ACLED), a disaggregated data collection, analysis, and crisis mapping project, maintains a database of all forms of human conflict from over 50 developing countries.ACLED is the most widely used real-time data and analysis source on political violence and protest around the world. It collects the dates, actors, locations, fatalities, and modalities of all reported political violence and protest events across major regions, including Africa, South and Southeast Asia, Central Asia and the Caucasus, the Middle East, Latin America, the Caribbean, Southeastern and Eastern Europe, and the Balkans. ACLED uses four types of data sources for its analysis: traditional media, reports from NGOs/governments, local partner data, and social media. Each week, ACLED researchers analyze thousands of sources in multiple languages to provide the most comprehensive database on political violence and demonstrations.

this graph was created in R :

https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F16731800%2F418952a3857f2530a53a40d9cc9c320c%2Fgraph1.gif?generation=1732477206118972&alt=media" alt="">

Due to the size of the full dataset (see Technical Notices below for more information), users are advised to download data for specific time periods and/or geographic areas.

To download all available ACLED data for a specific time period, enter your login information, select a date range in the ‘from’ and ‘to’ boxes, and click ‘export.’ To download all available ACLED data for a specific region, country, or location enter your login information, select a ‘region,’ ‘country,’ or ‘location’ from the relevant drop-down menus, and click ‘export.’ Note: ‘country’ selection will override ‘region’ selection, and only data for the selected country or countries will be downloaded. ‘Location’ selection requires a ‘country’ selection, and will result in an export of only data for that specific subnational location.

To download data for specific event types, select the relevant event types from that category in the ‘event type’ or ‘sub-event type’ boxes and leave all other categories as they are. All data for the selected event type(s) will be exported.

To download data for a specific actor type or a specific actor, select the ‘actor type’ or ‘actor’ in the relevant boxes and leave all other categories as they are. All data for the selected actor or actor type(s) or actor will be exported.

By default, the data are exported in a format where each row represents a single event, on a specific day and location, and involving distinct actors. An ‘actor based’ file displays events by single actors instead, meaning that events are often repeated if two actors are involved. To determine which of the two file types to use, you should consider whether the data are being used to analyze patterns over time, types of violence, conflict between groups, or locations (which the default file type is best for), or to analyze actor types or specific actors. For the former, the default format should be used, while for the latter, the ‘actor based’ file should be used.

For systems that use semi-colon separated values by default, you may wish to use the ‘compatibility mode’ option.