Objective
The main objective of the project is to develop and validate a recommendation system for destination decision-making. It will provide personalized recommendations based on user profile and contextual information. It will be compliant with existing eCommerce and technological standards (component based and multi tiers) and give support for accessing heterogeneous data, from plain text to semi structured (XML) and structured (RDBMS). It will be interactive and "conversational" adapting the dialogue process as it learns more about the user. The dialogue will be driven by a tourist destination selection decision-making model. It will support product aggregation for a given destination in the form of a ""travel plan"" that can be used to support the user during the travel (recall events, advise on traffic condition, etc.) by means of his/her portable phone. The project will bring innovative results in the area of Case-Based Recommendation Systems and Tourist Decision Modelling.
Work description:
The project can be subdivided into four temporal phases: Development of the decision model. In this first stage we shall collect the requirements (use cases and system features). Moreover, previous available knowledge on user behaviour, in the form of cases (human/human and human/compute interaction logs), will be analysed. That data will enable to develop a realistic model of the user decision process, including the modelling of destination.
First prototype design and development.
The key technologies will be chosen and for some of them a quick evaluation will be conducted (XML data management system). This material and the decision model will provide input for the design and the development of the first prototype, a fully operational recommendation system to be integrated in two existing tourist portal (managed by two partners in the consortium).
The key components are:
- dialogue management based on the tourist decision model;
- similarity based queries;
- filtering using a user model;
- user activity logging.
Prototype management and evaluation.
We want to conduct rigorous statistical tests to determine the real significance of the improvements brought by each single technique. In this stage alternative approaches (for filtering, selecting, presentation), supported by the prototype, will be run in parallel and evaluated with users.
Final recommendation system and framework.
Implementation of a set of requirements not yet done, that is the personalized presentation of the destination and the Travel plan tool. A wizard that enables the user to store in a single entry point a collection of information connected to the travel. These data will help the service providers to follow and support the user during the travel. Moreover, at this stage, log data collected with the first prototype will be used for user profile learning and tuning the various filtering techniques implemented.
Milestones:
- M8: Tourist decision model: model of the decision variables and process.
- M9: Software Architecture: data integration and representation (customized views).
- M16: Case-Based Reasoning components: software library (cases in XML and retrieval functions).
- M18: First Prototype: selection of personalized destination; integration with existing data repositories.
- M29: Final system: personalized presentation of destination; content and collaboration based filtering; travel plan management.
Fields of science (EuroSciVoc)
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
- natural sciencescomputer and information sciencessoftware
- social scienceseconomics and businessbusiness and managementcommercee-commerce
You need to log in or register to use this function
Call for proposal
Data not availableFunding Scheme
CSC - Cost-sharing contractsCoordinator
38100 TRENTO
Italy