Semantic Web Aided Itinerary Planner Rohit Sud Aditya
Semantic Web Aided Itinerary Planner Rohit Sud Aditya Sakhuja Mayur Bhosle Aditya Devurkar Course: CS 8803 AIAD Prof: Ling Liu
Objectives Provide users with an itinerary that is aligned with their specific interests Automate the process
Background and motivation Complete Vacation Planners like : www. tourguidemike. com The results are not personalized. User plans his own itinerary: www. mapquest. com Hence… Need for a system which relieves the user from manually planning a trip Services like (www. wikitravel. org) provide excellent and extensive information about almost all places to visit
Putting them together: Research Problems An intelligent itinerary generator. The current systems are static in nature We propose an intelligent system which takes decisions inferred from the available knowledge – using user's profile and trip profile Semantic data Extraction – Major challenge due to area being unexplored. RDF – Querying RDF data sources – SPARQL Effectively identifying the relevant factors for the itinerary generator
System architecture
Algorithm Description Itinerary Plan Generator : Inputs Static data – User Profile (Interests) Trip data – Start/end date, starting location, “min cities to cover”, cost limits, activities , location category Distance Matrix Similarity Matrices Traveler – City Similarity Matrix City – City Similarity Matrix Location specific weather, traffic reports
Algorithm Description Itinerary Plan Generator where, wi are the weights, xi are the factors The factors for first level of pruning Interest – Category mapping Cost factor = ( Expected Cost – Limit ) Travelling time Location Hotness Rating = ∑ Ratings i / N ; where N - Users
Algorithm Description Itinerary Plan Generator Traveller-City similarity matrix Similarity equation factors Events : Activity mapping City category : User interest mapping Spot category : User interest mapping Calculation the score -
Algorithm Description: Feedback Is the user satisfied with the results ? Is the user happy with the weather and traffic updates ? Traffic Sunnyvale Weather Sunnyvale
RDF : Description RDF Extractor module RDF? Why RDF? SPARQL querying. City Name SPARQL Engine City Details RDF
Algorithm Description: RDF Our Sample Schema in RDF/XML. <? xml version="1. 0" encoding="UTF-8" ? > <rdf: RDF xmlns: rdf="http: //www. w 3. org/1999/02/22 -rdf-syntax-ns#" xmlns: ct="http: //city. org/Info#" xmlns: spot="http: //city. org/spot#"> <rdf: Description about=“Some City"> <ct: name>Atlanta</ct: name> <ct: latitude>Val_Latitude</ct: latitude> <ct: longitude>Val_Longitude</ct: longitude> <ct: incityspots> <rdf: Description about="s 1"> <spot: name>s 1</spot: name> <spot: famousfor rdf: resource="Athletics"/> <spot: famousfor rdf: resource="Trekking"/> <spot: bestvisittime rdf: resource="morning"/> <spot: spenttime rdf: resource="6 hr"/> <spot: costsyou rdf: resource="$5000"/> </rdf: Description> </ct: incityspots> </rdf: RDF>
Generating User Reports The result of the trip is published on a Google Map We use the Google Map JS API for it.
Using Google Maps API Clicking on maximize displays information about the place from RDF data source.
Evaluation of Results • “Leave one out” test to evaluate the correctness of the algorithm • “Freshness” of results is ensured by giving dynamic data to the user • “User Satisfaction” is measured using the feedback loop
References [1] http: //www. tourguidemike. com/ [2] http: //www. lonelyplanet. com/ [3] http: //www. mapquest. com [4] http: //www. tripit. com/ [5] "Crumpet: Creation of user-friendly mobile services personalised for tourism", Stfan Poslad, Heimo Laamanen, Rainer Malaka, Achim Nick, Phil Buckle and Alexander Zip [6]Cyberguide: A Mobile context aware tour guide. [7] http: //www. travelok. com/ [8] http: //wikitravel. org/en/Main_Page [9] http: //www. holidayandtravelguide. com/ [10] http: //www. holidaytraveldestinations. com
Questions?
- Slides: 16