Global Initiative of Academic Networks (GIAN) organized a short term course on Spatial Data Analytics at School of computer and Systems Science, JNU New Delhi from 26th Dec. - 30th Dec 2016. Faculty members Ms. Anu Rani and Ms. Pooja Dang from IT & CSIT department of Dronacharya College of Engineering, Gurgaon attended the short term course. 44 participants from the different technical colleges participated in course. The course was inaugurated by Prof. B. S. Butola, Chairperson CSRD, JNU. The Key Speaker of the course was Prof. Shashi Shekhar McKnight Distinguished University Professor, Department of Computer Science, and University of Minnesota, USA. The course was coordinated by Prof. Sona Jharia Minz and Dr. Ayesha Chaudhary, SCSS , JNU New Delhi.

Day1: 26th December 2016

Content of the day: Introduction and motivation on Spatial Data Analytics

The session started with Introduction and need of Spatial Data Analytics. Prof. Shashi Shekhar He explained the concept of Spatial data analytics i.e concerned with analysis of data describing geographic phenomena (e.g., climate) or instrumented physical environment (e.g., roads, building). It is important for societal applications in sustainable development, energy, mobility, public safety, public health, as well as emerging location - based services such as local advertisement and mobile commerce (e.g., Uber).

The course introduces the fundamental ideas and research challenges underlying the emerging spatial data analytics systems for spatial database management, spatial data mining, and spatial network engines underlying location based services.

Day 2: 27th December 2016

Content of the day: Spatial Query Language and Spatial Data Models

Dr. Shashi Shekhar, Professor, University of Minnesota started the lecture with motivation for spatial query language in spatial data analytics. He delivered the brief review about structure query language of databases. Then he extended the lecture to SQL / OGIS for spatial data. In this he discussed various syntaxes of SQL related to spatial data and use of OGIS software. In second half session Dr. Shashi Shekhar, delivered the lecture on Spatial Data Models. He related many real time problems with Spatial Data Models. In tutorial session participants got exercise and quiz sheets on spatial query language. At the end of the session feedback was taken.

Day 3: 28th December 2016

Content of the day: Spatial network and Location Based Services

Dr. Shashi Shekhar, Professor, University of Minnesota started the lecture with the use of spatial networks and various examples. He also discussed Conceptual model, Connect statement, Recursive statement and various data structures for spatial data. In second half session Dr. Shashi Shekhar, discussed the various algorithms for connectivity query, Dijkstra’s algorithm, A* algorithm and Hierarchical routing algorithm for shortest path. In tutorial session participants got tutorial and quiz sheet on spatial networks.

Day 4: 29th December 2016

Content of the day: Identifying patterns in spatial information

Prof. Shashi Shekhar started the session on Identifying patterns in spatial information. The explosive growth of spatial data and widespread use of spatial databases emphasize the need for the automated discovery of spatial knowledge. Spatial data mining is the process of discovering interesting and previously unknown, but potentially useful patterns from spatial databases. The complexity of spatial data and intrinsic spatial relationships limits the usefulness of conventional data mining techniques for extracting spatial patterns. Efficient tools for extracting information from geo - spatial data are crucial to organizations which make decisions based on large spatial datasets.

The next session was on computational issues: The volume of data, the complexity of spatial data types and relationships, and the need to identify spatial autocorrelation pose numerous computational challenges to the SDM field. When designing SDM algorithms, one has to take into account considerations such as space partitioning, predicate approximation, multidimensional data structures, etc. Table 2 summarizes how these requirements are in contrast with with classical data mining. Computational issues may arise due to high dimensionality of the spatial data set, spatial join process required in co-location mining and spatial outlier detection, estimation of SAR model parameters in the presence of large neighborhood matrix W, etc.

Day 5: 30th December 2016

Content of the day: Computing at the Nexus of Food, Energy, and Water

The topic of discussion was Computing at the Nexus of Food, Energy, and Water. He discussed that in coming decades, the world population is projected to grow significantly increasing the demand for food, water, energy, and other resources. Furthermore, these resource challenges may be amplified due to climate change and urbanization. In addition, Food, energy and water (FEW) systems were traditionally analyzed and planned independently to address the challenges of population growth, climate change and urbanization. However, such piece-meal approaches (e.g., bio-fuel subsidy, fertilizers in agriculture) to solving problems in one system (e.g., energy, food) led to unanticipated harms to other systems (e.g., food price increase, water resource depletion and degradation). Thus, understanding the interdependent and interconnected nature of food, energy, and water systems (FEW nexus) is a societal priority.

However, the FEW nexus presents new challenges and opportunities for computing. For example, data science methods need to not only re-examine assumptions such as non-stationarity (e.g., climate change) but also address nexus challenges such as high cost of false positives, (social) feedback loops, and multiple spatio-temporal scale.

Prof. Shashi Shekhar concluded the session by discussing the future perspectives of Spatial Data Network Analytics.

The course concluded with a panel session where Prof. Vishwanath Gunturi, IIIT Delhi , Prof . Ayesha Chaudhary, and Prof. Sona Jharia Miniz JNU Delhi shared their views and gave the vote of thanks. It was a very knowledgeable session and as the participants learnt a lot about Spatial Data Analysis.