Currently, my broad research and technology interests include
- Graph matching algorithms and spatial alignment .
- Peer-to-peer systems and distributed consensus with a bias towards geospatial applications .
- Modeling community, customary, and indigenous land tenure systems.
- Distributed land information systems as technology solutions for customary or indigenous land tenure regimes .
During my PhD years, my research is focussed on application Qualitative Spatial Knowledge Representation and Reasoning to spatial map alignment. Spatial map alignment is the discovery of correspondences between a pair of maps (e.g. the map of a neighborhood in a city and the map of whole city). Map alignment can be used as a technique for visual query answering. It can also be used in applications that require spatial localization at a scene level but without the use of precise coordinates. We have applied spatial alignment methods in our projects sketchmapia and its4land.
The special aspect of my work is that I consider spatial information given purely in terms of qualitative relations. In particular, I have developed several metaheuristic algorithms for solving the map alignment problem by matching qualitative spatial representations of the input maps (see my publications below).
From this initial pursuit my research has taken me into several domains including algorithm analysis and design, optimization, computer vision (semantic sketch understanding), domain modeling and ontology development (using owl and answer set programming to specify and query customary land tenure domain models), and distributed peer-to-peer systems (implementing and deploying user controlled systems with applications to community land registers and peer-to-peer markets).