- Casts Creation
- Video annotation
- Casts vs. Projects
- Cast Space Browsing
- Casts Popularity Assessment
- Casts Popularity Visualization
- Content Filtering / Tagging System
- Reputation System
Casts are what we define as location-based and user-generated content within Locast. They consist of multimedia (video or audio) containing a geo-coordinate, a description and geo-semantical tags that can be both user-generated or recommended by the platform.
Casts can be generated in a number of different ways. The Locast mobile application for Android-enabled devices allows you to create casts and share them right away on the website. If you don’t own this kind of handset, you can still actively participate to Locast by producing your own casts. You can use any digital device: a mobile with built-in camera, a photocamera, a laptop with webcam or a digital audio recorder.
Locast mobile application allows to quickly annotate a video being just recorded adding cue-points in the frame sequence and personal comments that can be useful during fine editing operations on the content. Video annotation tool allows also to rapidly edit the video taken, re-allocating entire sequences or adding casts being imported from Locast database
When creating content you have two main options. The first is called a Cast, this is the more free-form and spontaneous option. A cast is basically when you use your mobile to shoot a video (or audionote) that gets automatically georeferenced. The second option is more structured, and we call it a Project. Shortly, a project is a group of casts related to a topic / urban area being produced by a group of reporters.Cast / Report
A cast / report is a common option that consists of creating an individual cast on a topic or event happening in a specific location. These actions are not necessarily planned in advance and are normally carried out by a single person.Project
Projects are scheduled actions carried out by a group of citizen journalists. They can be created according to a specific location of the city, a specific topic, or both. Projects are like labeled containers in which multiple casts get associated according to a common place or topic. Topic coverages can be either declared as opened - which means everybody would be able to join them - or closed - the participants are pre-defined by the coverage creator(s).
Casts can be easily accessed and watched on the Locast website. Casts might also be watched on a mobile phone. The Locast application installed on an Android-enabled phone can detect where you are in the physical space and suggest casts related to the nearby area.
Casts can enable conversations among Locast community members. The platform provides the possibility to directly reply to each cast through text-based comments or video-based ones. Private messaging among users is also supported.
Locast staff strongly encourages citizen journalists to be accountable for the content they create and share on the platform - as a form of democracy and transparency. At the same time, we are aware that covering certain topics can be a dicey prospect for the producer of that content. Accordingly, when sharing a new cast you have the possibility to anonymize the process - so that the content will appear on the website without personal references. However, casts uploaded as anonymous will be not available right away, but placed in quarantine to be approved later by trusted editors. It is worth remarking, that although casts appear to users as anonymous, our trusted editors potentially have access to authorship information.
Following the Wiki paradigm, Locast contains a versioning system that allows you to see the evolution of a certain cast. Such system is an ideal tool for citizen journalists that want to edit and optimize their own casts that were produced in the streets.
Locast keeps track of each edit and saves each cast’s history so that it can be browsed in a transparent way by the entire community.
Locast editors are trusted users and staff members that perform editing activities on casts after they have been posted. In fact, they don’t censor or perform any other content restriction on casts being uploaded. Their participation only occurs whenever:
- a cast is flagged as inappropriate
- a cast is uploaded as anonymous
Editors also have the ability to feature worthy casts on the Locast - Civic Media headline in the homepage. This is normally done to give higher visibility to outstanding content.
The overall popularity of a cast is decided on a weighted combination of the following datapoints: cast views, times favorited, number of likes / don’t likes, number of comments, and times downloaded. These datapoints are combined with the time parameter in order to generate the heat visualization proposed on Locast Civic Media.
A heat map of colors represents the level of popularity of a casts within the website. For instance, a bright red represents the newest, most popular content. As the content is consumed over time, it loses its novelty value and “cools off”. This heat color has the possibility of rapidly changing in response to recent activities on the part of Locast users.
As a cast moves throughout its life, it may collect artifacts of how others were affected by it (views, comments, etc). To represent this collection, we are proposing to use a border of various thicknesses. The thickness of the border would only be able to increase in response to all user activities over the life of the cast. Thickness of lines would vary between one and ten pixels.
Cast is first posted:
- Coolest color, smallest border.
- A few people see the cast, and post it to other social networking sites:
- Color heats up, but not enough activity to increase border size.
Cast is getting more hits, starting to get popular:
- Color is much warmer and border increased.
Cast is spreading:
- Color heat jumps, as does border thickness.
A week later, there are still a few people watching, but much less than before. Cast cools off, but border continues to grow.
Several months later, the novelty has worn off, and the cast is barely viewed at all. Cast continues to cool, as border slowly grows.
The sheer number of casts within Locast can be overwhelming. In order to avoid information overload issues and to allow users to tailor the content listing to the personal preferences, a content filtering based on tags gives the possibility to sort and prioritize casts. Locast tags are both related to semantical aspects (typology of content) and location-based ones (geo-coordinates). As to the filters, these are based on a tri-state modality:
- Preferred: casts described with selected geo-semantical tags having high-priority for a certain user.
- Neutral: content tags that are not preferred nor ignored. In the listings and visualizations, related casts would appear right after the preferred ones.
- Ignored: tags defined as ignored would make related casts disappearing from the lists.
A reputation system is a relevant part in platforms where the contents is generated by users that don’t necessarily know each other. Furthermore, a reputation system is an effective way to assess the participation of each Locast member and its goal is to assign priorities to those participants whose contributions are considered as more relevant and high-quality.
Locast implements a reputation algorithm that measure the personal development of each member within the community assigning him a qualitative ”label” which value is considered proportionally according to the other members’ performances.
Locast evaluates a number of positively and negatively evaluated actions in the assignment of the labels. These actions correspond to:
- each log-in session (3 points)
- each cast uploaded (30 points)
- each revision to casts (10 points)
- each comment (10 points)
- creation / participation to an assignment with at least 2 members and 4 casts associated to (20 points)
- flagged & censored casts (-100 points)
Named levels within Locast reputation system are:
Users that recently subscribed and/or are not active in the content production -community conversation processes.
For community members with no contributions in terms of casts but active in the conversations.
Members with uploaded casts and a fair participation to the community.
Members with a remarkable participation in terms of uploaded casts, participation to the community and projects.
Members with outstanding community participation and contributions.