Wednesday, December 9, 2009

End of the Quarter Colloquium for Fall 2009



In the last year, several events have been held to try to create an atmosphere of networking in the technology sector. Margaritas at the Matador, after work beers sponsored by the WTIA at Meconi's, and ironically the one event with no drinking called "Barcamp" at Suite133. We just came off of a very successful South Sound Tech Conference as well.

What may be overlooked, in regards to keeping up with what is happening in technology on an ongoing basis, is the end of the quarter colloquiums at the Institute of Technology. Students at the undergraduate and graduate level participate in a day of presentations regarding internships, readings and research. Those of you who have interned students may have already attended one. If there is a subject area you are interested in, you are welcome to attend and network as well.

This Friday is the colloquium for Fall 2009. Here is some info on it.

----------------------------------------------

Institute of Technology Student Colloquium

Fall Quarter 2009

Speaker Session One (Cherr Parkes Building, room 108):

09:30am Spatial-Temporal Access Control for e-Health Services
Apaporn Boonyarattaphan (Dr. Yan Bai (Chair), Dr. Sam Chung, Dr. Ankur Teredesai, Dr. Radha Poovendran)

The transformation of healthcare from human-based to online services can expose e-health to the security threats as other online applications. The identities of legitimate e-health users need to be verified cautiously before the access is granted. Given that any healthcare services only happen at certain times and certain locations, we propose use time and location to perform security control. In particular, we will develop and implement a prototype of the Spatial-Temporal Access Control (STAC) for e-health services to authenticate and authorize users. The traditional mode, Role-Based Access Control (RBAC), is too rigid to handle the dynamic authentication and authorization requirements of e-health applications. STAC is beneficial for e-health services since it allows us to define the spatial and temporal constraints for e-health authentication and authorization decisions.


10:00am Finding the Best Parser for PRISE - a Comparative Study
Craig Truzzi (Dr. Martine De Cock, Dr. Ankur Teredesai, Drs. Timur Fayruzov)

The Protein Interaction Search Engine (PRISE) is a biomedical text mining system intended to help researchers find interactions between proteins described in full text. One of the cornerstones of PRISE is an English language parser. The parser is essential for the text mining process, as it converts a plain English text sentence into a set of features, such as part-of-speech (POS tags) and grammatical dependencies, which are used to extract interactions. In the current PRISE implementation, the Stanford Parser is used to process the text. It is a high quality parser that implements some interesting features, but it is too slow and allows PRISE to process only a couple of thousands of abstracts per day.

In this presentation we will explain how a study of the (dis)advantages of various other parsers as documented in the literature led us to propose the parser NO-RERANK as an alternative for the Stanford parser in PRISE. We have implemented this other parser into PRISE to empirically evaluate its impact on the system. In this talk we will reveal whether the replacement gave rise to an increase in efficiency and effectiveness and, if so, to what extent.


10:30am Ranking of Protein Interactions for PRISE
George Dittmar (Dr. Martine De Cock, Dr. Ankur Teredesai, Drs. Timur Fayruzov)

The Protein Interaction Search Engine (PRISE) is a biomedical text mining system intended to help researchers find interactions between proteins described in full text. This presentation will discuss our work towards developing a new ranking module for PRISE to help rank abstracts returned to the protein researcher. The main objectives of our project were to evaluate current ranking and output schemes used by PRISE and test them against newly proposed ranking strategies. We will discuss the implementation of different ranking schemes over different annotated biomedical data sets, as well as present precison@k values and what they mean in terms of the results obtained doing test runs of the different schemes.


11:00am Twitter Authority Based Search
Rinkesh Nagmoti (Dr. Ankur Teredasai, Dr.Martine De Cock)

Microblogging services such as Twitter are increasingly becoming a valuable source of information and are therefore gaining tremendous importance for web search and real time data analysis. In this project we propose Authority Ranking algorithms for Twitter users to improve Information Retrieval and Search for Twitter. We also demonstrate a web based tool which allows users who are interested in searching Twitter, to compare various search results. The demonstrated tool includes a twitter search engine interface that comparatively presents relevant tweets to the end-user. The end-user then provides her preference for one search result over another; and the tool records this preference and reports on overall search statistics. These statistics helps in analyzing the Authority Ranking Algorithms which in turn helps in improving algorithms which powers the real time microblog search.


11:30am A Working Prototype of a Daily Dietary Social Network
Billy Lybyer, Andrew Park, James Tolman, Jung Shin
Dr. Ankur Teredesai)

The social web is becoming increasingly pervasive with the advent of mobile interfaces that allow real-time, on the fly publish-subscribe mechanisms. Given the advances in this domain, aggregating information about what one eats on a daily basis should not be difficult, yet collecting, querying and sharing this data with healthcare professionals such as doctors and nutritionists is still a challenging task. Integrating such data with other electronic medical records data systems may lead to significant benefits for not only the subscriber but the overall society. Many restaurants now provide information about the contents of their menus online and there is avid interest amongst users to collect and share this information as effortlessly as possible.


The focus of this project is to develop a working prototype of a social network whose main intent is the sharing of daily dietary food intake amongst its members. The main access point to the network would be by the use of an intelligent mobile application. The members of the social network would be able to take photographs of the food they eat and add textual ‘tags’ to their photos using their phone. An additional option is to allow just textual information about their food intake not associated with any photos.

12:00pm Computer Cooking Contest 2009 Website
Wilfredo R. Ortiz (Dr. Isabelle Bichindaritz)

This presentation explains the difficulties and actions taken to update a web application written for the Computer Cooking Contest 2009. This contest involves creating a website that can be used to research and search for recipes based on certain user-entered criteria. The portion I dealt with was a restructuring of the web application layout, including server configuration, database interactions and the user interface (web design).

This project required expanding on the original project's design, and implementing cleaner user interfaces for the searches as well as modifying how search results were parsed and displayed. These now use JSPs, Javabeans and JSTL tags for easier and more configurable display and greater customization and standardization across pages as opposed to the previous HTML pages. The difficulties in deploying a Netbeans project to an external Tomcat server served as a valuable lesson, including how to implement Catalina to run Java servlets.

This project matched closely with my current internship and allowed me to expand my knowledge of the techniques and software that were used there. In fact, by completing this project I was able to work more effectively due to a better understanding of the frameworks and technologies involved. This was especially true in the case of Javabeans and the JSTL library, which I had little experience using beforehand. Overall, the project did not hinder but in fact enhanced my internship experience, and is an example of how classes at the University of Washington Tacoma can prepare you for your chosen career field.

12:30pm Avue Technologies Internship
Wilfredo R. Ortiz (Dr. Isabelle Bichindaritz)

This presentation will describe my experiences as a Software Development Intern in a professional field working at Avue Technologies. As a first foray into the field for a full-time student it is important to know what technologies and techniques are currently in use, and how the time spent and information acquired at the University of Washington Tacoma's Institute of Technology ties into a career in this area. That is the purpose of this presentation.

As far as relevant technologies go there are several that were familiar and several that took time to learn. New to me was the Spring Framework, a Java framework for building applications, and in this case web applications. Also presented was Enterprise JavaBeans Technology, another architecture for building web applications. Furthermore, I gained experience concerning database/website interaction and an introduction to several development tools that I had never worked with before this internship.

Overall my time spent at the university did assist in preparing me for entering this career field. The intensity and magnitude of the work involved in the professional sector served as the greatest difference from the University environment. Looking back at my time spent at Avue Technologies this quarter, I now see why an internship is recommended for this degree choice, because without it you will have little to no idea what you will be dealing with when you enter the job market.

01:00pm Research In Motion: Software Development in a Mobile World
Tim Simon (Dr. Daniel Zimmerman)

As cellular phones continue to increase in power and popularity, the mobile development market also continues to grow and change. It is predicted that in 2014, 6.67 Billion mobile applications will be downloaded. In 2009, the Blackberry RIM market share increased from 40.4% to 47.5% (IDC Worldwide Quarterly Mobile Phone Tracker, 2009). As mobile applications are significantly less complex than desktop programs, it is possible for a small group or even an individual to develop and market software targeted at the ultra-portable market.

This presentation will give a brief look at the mobile market, and then delve into techniques used for creating mobile applications in the Blackberry Research In Motion (RIM) environment. The discussion will cover the development environment, unit and black-box testing, style differences, and API, as well as difficulties that can be expected during the coding process. After this presentation, the viewers should have an idea of how to begin developing for the mobile environment in general, and the RIM platform specifically.


01:30pm From Design to Release - Game Development
John Emerson (Dr. Daniel Zimmerman)

A common adage in the game development industry is "The last 10% of the project is 90% of the work". I have found this to be true, so I intended to take a game project from the design stage to distribution stage, and overcoming all hurdles in the way. Not having succeeded, I'll examine why and present the progress that I've made.

02:00pm FPGA Design Engineer Intern
Daniel De Jager (Dr. Larry Wear)

Daniel De Jager discusses his experience working as an FPGA Design Engineer for EMAG Technologies, an Ann Arbor MI based company that specializes in original solutions to complex RF problems. In his work, he was tasked with designing a device that upon receiving a trigger signal will start creating a 2.5Gbit/sec signal of arbitrary content. This project allowed him first-hand experience of a practical application of FPGAs. During his stay at the company, he was able to practice design methodologies taught at UWT and make good use of those skills to go from concept to practical design and working simulations.

1 comment:

Anonymous said...

Can anyone recommend the best Managed Service tool for a small IT service company like mine? Does anyone use Kaseya.com or GFI.com? How do they compare to these guys I found recently: [url=http://www.n-able.com] N-able N-central service management
[/url] ? What is your best take in cost vs performance among those three? I need a good advice please... Thanks in advance!

Second Book of Short Stories Out on Amazon

During the last year I wrote four short read collections called the Three Twisted Tales . Today they are available in aggregate as The Clock...