Presentation of a Geoportal Based on Semantic Recommender Systems

These days Geospatial data has important roles in every organization and ordinary people’s life. They need these kinds of data for each activity. The most common believe in Geosciences is that which more than 80 percent every type of data in every organization has a spatial component which is produced by public sector, commercial companies and people (Volunteered Geospatial data). In addition to available data, development of data capturing technologies has led to increase data production ratio every day. Information overload still remain as a big challenge in current web and because of information overload, there is also big challenges in the world of geospatial information system. Data and services on the web are beneficial when users could discover them in an efficient manner. Difficulties such as ambiguity of keywords used for resource description by data providers and data users, different context of data providers and data producers, difficulties which users encountered dealing with spatial portals interfaces and spatial resources, number of clicks which a user required to navigate to a search panel, the variety of discoverability criteria’s on search panels, irregular descriptions of spatial resources or insufficient information provided on spatial resources which listed on result panel could cause users wasting lots of time for investigating interfaces and spatial resources which cause bewilderment to users of Geoportal in discovering geospatial resources. geoportal is kind of portal which mostly used for geospatial resource. Geoportals as a gateway for SDI is a tool which helps expert and non-expert users to discover geo information services. Geoportals have lots of functionalities like address identification, measuring, map viewing, printing etc. Most of the different approaches which have been utilized to resolve these problems emphasis on keyword based enhancement of data discoverability so the mentioned problems still exists. The aim of this research is to enhance Geoportals functionality by integrating new information retrieval procedure in Geoportals named as recommender systems. These days most of the web portals and information sources have a process of recommending items to users which could attract their attention and satisfies their expectations with personalizing their content for them. Personalized systems like recommender system assists users to find content, products, services (such as movies, music, news, articles .etc.) using processed and combined ratings (Like/Dislike) which collected from other users interactions on the website. Recommender systems are new paradigm in information discovering and presentation to the users and are a very different procedure from available data retrieval procedures. We offered a geoportal architecture which has an embedded recommender component.we aimed to integrate a recommender system in Geoportal. In this case we could help to conquer the existed problems in Geoportals by recommending related items to users based on user profiles, information item metadata and extracting their similarities. We applied ontological similarity as a semantic similarity tool for calculating relatedness of a geospatial resource metadata and a geoportals user. Our semantic recommender system uses ontological similarity and Singular Value Decomposition (SVD) to predicting related items and to overcome cold start problem. We developed a geoportal based on Road Management and Transportation Organization Geodatabase and designed ontology.