Shambhavi Dhargalkar

Over the years, healthcare industry has evolved into adopting technology as an integral part of their system. More and more healthcare facilities and medical applications are using cloud based services for storage of medical records of their patients and users. With the increase in use of 3rd party cloud services for storage, the security risks towards the stored data have risen. These medical records could be shared with unauthorized 3rd party for business purposes posing a threat to the… Show more

Over the years, healthcare industry has evolved into adopting technology as an integral part of their system. More and more healthcare facilities and medical applications are using cloud based services for storage of medical records of their patients and users. With the increase in use of 3rd party cloud services for storage, the security risks towards the stored data have risen. These medical records could be shared with unauthorized 3rd party for business purposes posing a threat to the privacy of the patients. Our approach to solve this problem is to use a combination of symmetric and asymmetric key encryption techniques to provide maximum security to the stored medical records. There are 3 functionalities that a healthcare application will support i.e. uploading medical records, sharing medical data (with doctors or nurses) and updating medical data. Our approach tries to cover the 2 primary rules of security that are confidentiality, integrity along with authentication for the above mentioned functions, where encryption techniques like AES256 encrypts data and keeps it confidential, SHA256 message digest provides integrity
and digital signature ensure authentication. Show less

This project is part of the larger CoBRA (Comic Book Readership Archive) project that focuses on documenting comic book readership and fan related materials, such as fan mail, fan club publications, publisher sponsored contests, and additional related records. Our team created a digital archive of Avengers and Fantastic Four fandom content. Additional materials from other publishers and comic books will be added to the archive in the future. The purpose of our project is to create a… Show more

This project is part of the larger CoBRA (Comic Book Readership Archive) project that focuses on documenting comic book readership and fan related materials, such as fan mail, fan club publications, publisher sponsored contests, and additional related records. Our team created a digital archive of Avengers and Fantastic Four fandom content. Additional materials from other publishers and comic books will be added to the archive in the future. The purpose of our project is to create a visualization of data for CoBRA. The visualization is intended to give users insight about comic book readership fandom materials from the years 1961-1973 for The Avengers and The Fantastic 4. We have three visualizations describing various insights. The first one, GIZMO (Geo Information via Zoomable Map Organization) maps fans to their locations; a zoomable interface allows drilling down to different hierarchies based on locations. The second one, ART (Author Records via Timeline) is a static visualization comparing fan page, fan club, and type of contributor data along with a timeline of key events. The timeline includes events that are related to the data such as dates of fan club inceptions and disbandments. The third one, GIXUS (Gender Information Across the United States) is a static visualization of gender distribution across the US.
Show less

The rule based production system for Medical domain took in rules and managed a working memory of facts. The rules were represented in the form of antecedents and conclusions. Based on the facts provided in the working memory, appropriate conclusions were given by our system. The system also has a question mode, where the users are prompted to enter information about the patients, which updates the working memory and based on the updated facts, gives the appropriate conclusions. The… Show more

The rule based production system for Medical domain took in rules and managed a working memory of facts. The rules were represented in the form of antecedents and conclusions. Based on the facts provided in the working memory, appropriate conclusions were given by our system. The system also has a question mode, where the users are prompted to enter information about the patients, which updates the working memory and based on the updated facts, gives the appropriate conclusions. The implementation was done using Python. Show less

The project involved working on the real world dataset provided by Yelp. Two main tasks consisted of predicting the category and the ratings of a particular business based on the reviews provided by the users.Unsupervised learning approach was adopted for predicting the category and sentiment analysis of the reviews helped in getting the ratings of the businesses.

Implemented a basic search engine in Java. The ranking function used in the algorithm was based on the calculations involving the Term Frequency (TF) and Inverse Document Frequency (IDF). Indexing was done using the Lucene API.

Designed and developed a web recommender system for small to medium level businesses. The system facilitates the user to create a new account and upload his/her business services as well as opt for a solution that has been recommended by a majority of users for a similar profile. The ratings for listed businesses are taken on a 1-5 scale. Item Based Collaborative Filtering Algorithm is used along with a Semantics Keyword Search approach to generate recommendations that are most appropriate for… Show more

Designed and developed a web recommender system for small to medium level businesses. The system facilitates the user to create a new account and upload his/her business services as well as opt for a solution that has been recommended by a majority of users for a similar profile. The ratings for listed businesses are taken on a 1-5 scale. Item Based Collaborative Filtering Algorithm is used along with a Semantics Keyword Search approach to generate recommendations that are most appropriate for the user requirements. Show less

The project included a comparative study of all the major security protocols (WEP, WPA, and WPA2) available in the market today. Also, the solution for the Hole 196 vulnerability found in WPA2 security protocol when sharing a Group Transient Key (GTK) to all the peers connected over a network was implemented.