Quality Health Information Retrieval Based on Collaborative Filtering and Friendsourcing
|
If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL). |
Abstract
Background
Website content quality is critical in the health domain. Users need to retrieve health information that is precise, believable, and relevant to their problem. With the exponential growth of web contents, quality-based Recommender Systems (RS) have become indispensable for discovering new information that might interest to web users.
Semantic Recommender Systems use ontologies to determine semantically similar items following the widely known content-based, context-aware recommendation techniques. Collaborative Filtering (CF) is well known as a technique to improve RS with a social feature. In traditional CF-based RS, entities are recommended to new users based on the stated preferences of other similar users. Research has shown than CF can be improved by using content features or hybrid content-collaborative features.
Quality tagging is the activity of tagging web content with quality assessments regarding, for example, credibility, believability, readability, timeliness, accessibility, etc. Collaboratively constructed quality evaluations can be used as an input for content recommendation.
Despite their success in the industry, RS suffer from several problems. For example, they ignore the connections among users, thus missing the opportunity to provide more accurate and personalized recommendations.
Objective
The goals of the QHIR (Quality based information retrieval) project is to adapt strategies for collaborative filtering and collaborative quality tagging, to produce recommendation of resources in the health care domain.
In terms of social web, CF falls in the category of crowdsourcing techniques, which use the wisdom of crowd’s theory. Crowdsourcing is a distributed problem-solving approach; in the classic use of the term, problems are broadcast to an unknown group of solvers in the form of an open call for solutions. Social tagging is a manner of crowdsourcing where the collaborative activity is to have a description of web contents. Research shows that, when dealing with friends in a social network, traditional crowdsourcing mechanisms struggle to motivate a large enough user population and to ensure accuracy of the collected information. Friendsourcing is a form of crowdsourcing aimed at collecting accurate information available only to a small, socially connected group of individuals.
The expected novel contribution of our project is to demonstrate that CF based on quality evaluations can improve the quality of the recommendations according to the nearest social context of the user. Specifically, in this research we propose to extend RS with a friendsourcing CF strategy, this means to calibrate quality-based CF algorithm with the social network quality tagging.
Results
We have successfully built a software platform that provides reusable recommendation services (web-services) based on collaborative quality tagging and friendsourcing strategies. It has been integrated into a social network site developed on the Elgg platform, which supports healthcare prevention and disease prevention programs at the Medical Service of University of Cauca in Colombia (http://esalud.unicauca.edu.co/redunisalud/). During the first half of 2012 evaluations will be conducted to identify correlation between the relevance of the recommendations, the type of friendsourcing strategy applied, and the quality attributes tagged by the users. Early results will be presented during the conference.
Website content quality is critical in the health domain. Users need to retrieve health information that is precise, believable, and relevant to their problem. With the exponential growth of web contents, quality-based Recommender Systems (RS) have become indispensable for discovering new information that might interest to web users.
Semantic Recommender Systems use ontologies to determine semantically similar items following the widely known content-based, context-aware recommendation techniques. Collaborative Filtering (CF) is well known as a technique to improve RS with a social feature. In traditional CF-based RS, entities are recommended to new users based on the stated preferences of other similar users. Research has shown than CF can be improved by using content features or hybrid content-collaborative features.
Quality tagging is the activity of tagging web content with quality assessments regarding, for example, credibility, believability, readability, timeliness, accessibility, etc. Collaboratively constructed quality evaluations can be used as an input for content recommendation.
Despite their success in the industry, RS suffer from several problems. For example, they ignore the connections among users, thus missing the opportunity to provide more accurate and personalized recommendations.
Objective
The goals of the QHIR (Quality based information retrieval) project is to adapt strategies for collaborative filtering and collaborative quality tagging, to produce recommendation of resources in the health care domain.
In terms of social web, CF falls in the category of crowdsourcing techniques, which use the wisdom of crowd’s theory. Crowdsourcing is a distributed problem-solving approach; in the classic use of the term, problems are broadcast to an unknown group of solvers in the form of an open call for solutions. Social tagging is a manner of crowdsourcing where the collaborative activity is to have a description of web contents. Research shows that, when dealing with friends in a social network, traditional crowdsourcing mechanisms struggle to motivate a large enough user population and to ensure accuracy of the collected information. Friendsourcing is a form of crowdsourcing aimed at collecting accurate information available only to a small, socially connected group of individuals.
The expected novel contribution of our project is to demonstrate that CF based on quality evaluations can improve the quality of the recommendations according to the nearest social context of the user. Specifically, in this research we propose to extend RS with a friendsourcing CF strategy, this means to calibrate quality-based CF algorithm with the social network quality tagging.
Results
We have successfully built a software platform that provides reusable recommendation services (web-services) based on collaborative quality tagging and friendsourcing strategies. It has been integrated into a social network site developed on the Elgg platform, which supports healthcare prevention and disease prevention programs at the Medical Service of University of Cauca in Colombia (http://esalud.unicauca.edu.co/redunisalud/). During the first half of 2012 evaluations will be conducted to identify correlation between the relevance of the recommendations, the type of friendsourcing strategy applied, and the quality attributes tagged by the users. Early results will be presented during the conference.
Medicine 2.0® is happy to support and promote other conferences and workshops in this area. Contact us to produce, disseminate and promote your conference or workshop under this label and in this event series. In addition, we are always looking for hosts of future World Congresses. Medicine 2.0® is a registered trademark of JMIR Publications Inc., the leading academic ehealth publisher.
This work is licensed under a Creative Commons Attribution 3.0 License.