PROJECT TITLE :

Anonymous Indexing of Health Conditions for a Similarity Measure

ABSTRACT :

A health social network is an online information service that facilitates info sharing between closely related members of a community with the identical or a similar health condition. Over the years, several automated recommender systems are developed for social NetWorking in order to assist users find their communities of interest. For health social networking, the best source of knowledge for measuring similarities of patients is that the medical info of the patients. But, it's not fascinating that such sensitive and non-public info be shared over the Net. This can be also true for many alternative security sensitive domains. A brand new data-sharing scheme is developed where each patient is represented as a tiny range of (presumably disjoint) d-words (discriminant words) and therefore the d-words are used to measure similarities between patients without revealing sensitive personal information. The d-words are straightforward words like “food,’’ and thus do not contain identifiable personal info. This makes our method an effective one-way hashing of patient assessments for a similarity live. The d-words can be easily shared on the Internet to find peers who may have similar health conditions.