|
Computer Science Papers and Proposals
|
| 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 |
| 2009-2010 Computer Science Papers and Proposals |
Interacting with Social Networks to Improve Health-aware Body Sensor Networks
David & Steven Bauschlicher
ABSTRACT: Over the last decade, the demand for efficient healthcare monitoring has increased and forced the health and wellness industry to embrace modern technological advances. Body Sensor Networks, or BSNs, can remotely collect patient data and upload vital statistics to servers over the Internet. Advances in wireless technologies such as cellular devices and Bluetooth increase the mobility patients experience while wearing a body sensor network.
When connected by the proper framework, BSNs can efficiently monitor and record data while minimizing the energy expenditure of nodes in the BSN. Social networking sites play a large role in the aggregation and sharing of data between many users. Connecting a BSN to a social network creates the unique ability to share health related data with other users through social interaction. In this paper, we propose to integrate social networks and BSNs to establish a community promoting well being and great social awareness.
|
Optimization of a Touch Screen Soft Keyboard for Speed and Accuracy
Richard Davison
ABSTRACT: Mobile devices are becoming more mainstream in the lives of many people today, whether it be for business, school, or general entertainment. However, one problem which has manifested itself because of the shrinking of many of these devices is that of the keyboard. While many devices feature hardware keyboards, a growing number are opting for software keyboards, some of which feature touch responses and others which do not, each with varying accuracy. In this research, we developed a new chording input method for use with two thumbs, along with a search algorithm to generate layouts for said input method, to look at mobile keyboards from a different point of view.
|
Automatic Discovery of Categories in Video
Chris Reynia
ABSTRACT: As we move further in to the information age, the problem is no longer the lack of
availability of information, but rather the ability to find the right piece of information in
the overabundance of everything available. This is made even more difficult when the
piece of information being sought is not text, but a more complicated form of media such
as an image or video. We propose a method to help find video based off similarity to
other video, as well as automatically categorize this video based off an archetypal
categorization, such as genre. We do this by leveraging existing algorithms, which find
interesting frames in video, and then doing low level analysis on each of these frames.
This allows us to utilize well researched areas of computer vision in order to expand on a
comparably sparse field.
|
Dynamically Recommending Design Patterns
Sarah Richardson
ABSTRACT: Recommendation Systems for Software Engineering are created for a variety of purposes, such as recommending sample code or to call attention to bad coding practices (code
smells). We have created a system to recommend the use of design patterns. Many program-
mers have knowledge of design patterns but when they have limited experience implementing
them or are rushed to complete a product, the use of a particular design pattern may not occur
to them. We have developed a tool to dynamically search for signs that a programmer would
bene%uFB01t by using a particular design pattern, and make the appropriate recommendations to
the programmer as he or she is developing code.
|
Artificial Intelligence and Natural Language Processing in a Chinese Language Learning Game
Matt Samuels
ABSTRACT: Mandarin has been classified among the most difficult languages for foreigners to
learn as a second language. Not only is the vocabulary considerably different from that
of English, but also the grammar and Chinese writing system are considerably
complex. For this project correct Chinese sentences were used to create questions that
aid the learning of Mandarin Chinese. Different types of questions were created and
artificial intelligence has been incorporated into dividing questions into levels of
difficulty based on character or overall sentence complexity. Expert systems, including
real world data, were used in order to analyze the quality of the questions created.
|
An Implementation of Hadoop Utilizing Content-centric Networking
John Sherwood
ABSTRACT: The current state of networking has its foundations in design decisions that were made to deal with the realities of networking in the 1960s. Since that time, the advent of the Internet and the World Wide Web have reshaped the forces under which networking operates by shifting the focus from establishing a connection between machines to securely acquiring pieces of information with less consideration to which machine provided those pieces. To address
these changes, researchers predict that the creation of a global Content-Centric Network will be necessary, and one of the first steps in its successful adoption is the implementation of a variety of existing applications. Among the most widely used types of network software are the cloud platforms used by Google, Yahoo!, Amazon, and others. We propose to revise the TCP/IP reliant Hadoop framework used by Yahoo! to function on a content-centric networking protocol developed by Xerox%u2019s PARC labs.
|
Genealogy Research on the Web Utilizing Organized Guidelines
Anna Sterling
ABSTRACT: The Internet has enabled the spread of information with the
worlds%u2019 collaborative efforts, yet genealogists still finely sift through
individual search engines and applications to find the data that
matches. This project is proposed to organize this process of data
matching by utilizing a system that guides the researcher in following
standards of genealogical research methodologies. The application will
enable the classifying and linking of information found and the
resource it was found at. The application will walk a user through what
research needs to be accomplished next along with how to accomplish
it. It will be integrated with user-friendly dynamic charts, graphs, and
lineage retrieval reporting schemas. The researcher will be able to
directly connect with available APIs. The storing of data will be done
in a database designed for searching the varying dates, names, and
locations in order to automate the process of linking individuals with
existing data sets.
|
Augmented Reality for Later Playback of Videos
Jordan Stratton
ABSTRACT: The popularity of Augmented Reality has recently exploded since mobile phones have
pability to create a great user experience, while at the same time show useful
n. Unfortunately, augmented reality applications that have been created today
face a problem in that the user cannot later playback a video they have taken in order to
easily find information about places they have been. We propose a method of aluser to take a video and choose whether or not to have augmented reality on wrecord. Then the user can later playback the video with augmented reality. Thishelpful for people who want to record a place they have been and later want to
|
A Memetic Algorithm for Automated Music Composition
Derek Wells
ABSTRACT: Music has been an integral part of society for hundreds of years. Since the advent and rise of computers and computer technologies, musicians have utilized these advances in their craft, be it in electronic music, or the notation of musical scores. The field of memetic computing is a new and developing one. Research is being done in applying memetic algorithms to real world
problems. This research will offer a novel approach for a memetic algorithm to produce quality musical compositions. At the time of writing this paper, there were no proposals found in any literature for the use of memetic algorithms to compose music. This paper presents the results of
the development of four parent selection methods, a fitness function, and two crossover
techniques. Also present are objective and subjective testing results of the effectiveness of the
algorithm using the different parent selection methods, fitness function weighting, and crossover
techniques. The test results also show how the memetic algorithm improves over the genetic
algorithm.
|
Using Fuzzy Inference to Improve TCP Congestion Control Over Wireless Networks
Matt Wozniak
ABSTRACT: While modern wireless networks have been in development for a couple of decades,
the Transmission Control Protocol (TCP) which runs over those networks has existed
since the mid 1970s. As it was developed before wireless networks were even conceived, TCP was not optimized to consider the physical characteristics of wireless network links. Specifically, TCP responds to packet loss due to link errors in the same way it responds to packet loss due to congestion: it cuts back the rate at which traffic is sent. A means of improving performance of TCP over wireless links is to classify packet losses, and react only to those losses perceived as being caused by network congestion. This paper seeks to use environmental variables available to TCP implementations to feed a fuzzy inference system that will classify packet loss as due to congestion or random collision.
|
|
