- (Funded by Technology Innovation Hub, IIIT Delhi)
- PI: Anirban Mondal (Ashoka University, India); Co-PIs: Mukesh Mohania (IIIT Delhi, India) and Ladjel Bellatreche (ENSMA, Poitiers, France)
COLOURS is a personalized cognitive computing and social sensing based tourist assistance information system for creating a highly enhanced and immersive tourism experience in India. COLOURS collects hyperlocal information from local people, experts and historical/cultural documents to build a cognitive knowledge base. The objective of COLOURS is to design an effective user preference-aware and event-conscious navigation system for augmenting road trip planning in order to enrich user experience during the course of his/her journey. In particular, the objective is to provide the user with information concerning the positive and negative aspects/events pertaining to multiple routes corresponding to a given source and destination. Furthermore, we also aim at educating the users about their respective travel route(s) by informing them about historical/cultural facts about POIs in the routes (as well as the routes themselves) by means of storytelling. The architecture of COLOURS is provided below.
Notably, our problem entails several important research challenges, which we plan to address during the course of the project. In particular, it requires the design of novel algorithms and techniques for handling spatial queries with directives concerning POIs/events. Given that the directives associated with the queries may be abstract (e.g., trying to determine a “safe” route), this also requires investigation about how to quantify abstract notions such as “safety” in order to incorporate such notions into users’ trip planning. Another important research issue is about how to improve the efficiency of spatial query processing with the constraints associated with the query directives as well as the users’ event-oriented preferences. This requires the investigation of methods for caching/prefetching of the event data associated with the POIs for optimizing the search and communication costs. This also requires research into the reuse of the intermediate results of POI/event queries (with directives) across the journeys of multiple users. Furthermore, there are interesting research challenges concerning storytelling about POIs at multiple granularities.