Aleph’s data science research and development service takes initial ideas that our clients might have for data tools of perceived value to their organisation, and transforms them into a conceptual blueprint which captures in detail how the tool would achieve its aims.
This starts with defining the use cases for the proposed tool, capturing the tasks that would need to be performed by users, and its information inputs and analytical outputs. These use cases are then used to identify and articulate the functionality which is required of the tool.
These requirements are then translated into a set of operational components to be delivered through the application of specific data science approaches, with an explanation of which approaches are likely to be conceptually and practically optimal. The final conceptual blueprint brings all of these approaches together in the form of an integrated theoretical framework for how the tool is going to work.
Case Studies
01
Working on behalf of a UK university, Aleph produced a conceptual blueprint for the development of a claim verification tool.
This blueprint laid out what each component of the tool would need to do in order to extract claims from sources and assess their credibility, providing the entire theoretical basis for its operation and identifying a shortlist of candidate data science techniques that could be used. The accompanying report was used to seek additional funding to invest in the tool’s development.
02
We designed and documented a novel statistical approach for a law enforcement client to use in its assessment of the vulnerability of different locations to security risks.
This research considered available sources of data and the legal requirements associated with reporting site-based risks to formulate the modelling approach. It produced a viable conceptual framework on which to base the subsequent development of risk mapping tools for use in policing.
03
Aleph worked with a cross-government taskforce to examine the problem of keeping children safe online.
We were specifically charged with investigating ways that new technologies and emerging techniques might be combined to increase the certainty with which the age of a platform user might be verified. We designed an integrated method for using diverse indicators of age to provide probabilistic judgements about the age ranges of users. This method was used as part of the government’s engagement with technology companies on this issue.