4/28/17
Final Presentation Link
As the final week winds down of the senior project, I have begun to recap what I have learned over the past ten weeks.
Big Data is a powerful new way to handle information that will lead the future of technology and analytics. I truly believe we can do great things by processing larger and larger amounts of data, whether we begin to run models of the universe or simple human populations.
Hopefully this final presentation will shed some light on what I have discovered over the past 2 months. Thank you for reading!
Friday, April 28, 2017
Thursday, April 27, 2017
Final Thoughts
04/27/17
My time at Dignity Health has taught me the values of working within a team environment in the context of developing applications to big data. I have discovered what creative ways code and big data are being applied to handle hospital data and augment what the doctors would traditionally do on their own. Whether it be creating an alert system for sepsis, creating a genomics database, or just making an application that rates doctors based on their length of stay, technology has clearly revolutionized the way we handle healthcare.
Now, we can create large scale monitoring systems, systems to monitor those systems and so on. It all becomes increasingly complex as time goes on, but so does it in effectiveness of treating patients. Never before has a computer been able to take the load off of a doctor in terms of doing calculations or constantly monitoring different levels or values. It is truly a revolutionary feat. I have a newfound interest in what we can do with large-scale data collection, whether it be for medicine or simple consumer understanding in a more business focused scenario.
This project has given me a new mindset as to what I want to do in the future, I enjoyed working in healthcare more than I had anticipated, but it also reminded me why I wanted to work in something more consumer-facing. I enjoy creating things people will use, perfecting them for usability by your average person, whether it be hardware or software.
My time at Dignity Health has taught me the values of working within a team environment in the context of developing applications to big data. I have discovered what creative ways code and big data are being applied to handle hospital data and augment what the doctors would traditionally do on their own. Whether it be creating an alert system for sepsis, creating a genomics database, or just making an application that rates doctors based on their length of stay, technology has clearly revolutionized the way we handle healthcare.
Now, we can create large scale monitoring systems, systems to monitor those systems and so on. It all becomes increasingly complex as time goes on, but so does it in effectiveness of treating patients. Never before has a computer been able to take the load off of a doctor in terms of doing calculations or constantly monitoring different levels or values. It is truly a revolutionary feat. I have a newfound interest in what we can do with large-scale data collection, whether it be for medicine or simple consumer understanding in a more business focused scenario.
This project has given me a new mindset as to what I want to do in the future, I enjoyed working in healthcare more than I had anticipated, but it also reminded me why I wanted to work in something more consumer-facing. I enjoy creating things people will use, perfecting them for usability by your average person, whether it be hardware or software.
Sunday, April 23, 2017
Twitter Application Demo
04/23/17
Here is a demo of my Twitter data collection application that I have created and built over the time at my internship.
Image 1:
The code and program itself. (error was caused by killing the program just before capturing the image)
Image 2: I added a pop-up window to allow you to type in keywords, even multiple if they are separated by commas.
Image 5: It then writes to the CSV file I designated in the code
This is how I have gathered data to be analyzed through sentiment analysis to understand overall opinion towards a certain topic. Images in this post feature part of my collection of data for the term "healthcare".
Here is a demo of my Twitter data collection application that I have created and built over the time at my internship.
Image 1:
The code and program itself. (error was caused by killing the program just before capturing the image)
Image 2: I added a pop-up window to allow you to type in keywords, even multiple if they are separated by commas.
Image 3: Input the time to run.
Image 4: It begins to read the Twitter feed and then output it to the console below in Eclipse.
Image 5: It then writes to the CSV file I designated in the code
This is how I have gathered data to be analyzed through sentiment analysis to understand overall opinion towards a certain topic. Images in this post feature part of my collection of data for the term "healthcare".
Saturday, April 22, 2017
Twitter Analytics UPDATE
04/22/17
As I have progressed with many different projects over the past month, one that fell a bit by the wayside was my Twitter analytics program. Now, I have begun to kick it into high gear removing as many bugs as I can and essentially creating a working prototype of what we can do with twitter tracking. I can read the Twitter feed and write ALL the data to a .csv file which I can then import into an external program in SAS that can provide word clouds and the same sentiment analysis I was aiming to get through Google. I am attempting to secure a Google cloud account as to implement their sentiment analysis into the program itself to make it all seamless, but I am not able to do it at my internship due to their restrictions on personal account use. At the moment, I am refining it to make sure I can produce the results that can highlight the power of sentiment analysis on social media. I will be able to include some of my results in my final presentation, if anything using the SAS visual analytics tool to create some visual word clouds and present the sentiment analysis.
I am proud of what I was able to do and hopefully I can create a demo in a future post.
As I have progressed with many different projects over the past month, one that fell a bit by the wayside was my Twitter analytics program. Now, I have begun to kick it into high gear removing as many bugs as I can and essentially creating a working prototype of what we can do with twitter tracking. I can read the Twitter feed and write ALL the data to a .csv file which I can then import into an external program in SAS that can provide word clouds and the same sentiment analysis I was aiming to get through Google. I am attempting to secure a Google cloud account as to implement their sentiment analysis into the program itself to make it all seamless, but I am not able to do it at my internship due to their restrictions on personal account use. At the moment, I am refining it to make sure I can produce the results that can highlight the power of sentiment analysis on social media. I will be able to include some of my results in my final presentation, if anything using the SAS visual analytics tool to create some visual word clouds and present the sentiment analysis.
I am proud of what I was able to do and hopefully I can create a demo in a future post.
Sunday, April 16, 2017
Drug Price/Efficiency Dashboard
04/16/17
Hey Everyone,
This week I was introduced to a new project Dignity is working on that ties directly into what I had been doing with the prescription drug prices project. Essentially they are developing a system that can log and compare drugs based on their price and efficiency to the final patient as a means of making sure we are using the best value drugs, a balance of cost and quality.
I have largely been working on the actual design of the web application and making sure everything is functional, rather than developing the internal system as the groundwork is largely laid for that aspect. It is another instance of how they are gathering all the data surrounding outcomes of certain drugs and comparing it to the costs of each, exactly in line with what I had been researching in my previous efforts to ensure we were receiving the best prices compared to local pharmacies.
I wanted to share this with you as I am excited to continue my work with big data, thank you!
Hey Everyone,
This week I was introduced to a new project Dignity is working on that ties directly into what I had been doing with the prescription drug prices project. Essentially they are developing a system that can log and compare drugs based on their price and efficiency to the final patient as a means of making sure we are using the best value drugs, a balance of cost and quality.
I have largely been working on the actual design of the web application and making sure everything is functional, rather than developing the internal system as the groundwork is largely laid for that aspect. It is another instance of how they are gathering all the data surrounding outcomes of certain drugs and comparing it to the costs of each, exactly in line with what I had been researching in my previous efforts to ensure we were receiving the best prices compared to local pharmacies.
I wanted to share this with you as I am excited to continue my work with big data, thank you!
Saturday, April 15, 2017
IceScrum
04/15/17
In my previous post, I discussed many of the issues and complications that come with working in teams to create an application that may not even fit the end user's needs. It is a huge problem in any environment where the developers are not the people who will be using the application (which is nearly all development environments).
This week they began to roll out a new tool called "Icescrum". It is a form of scrum management that essentially provides a new way to organize the creation of an application.
Essentially, it begins with "user stories" that are requests or "stories" straight from the people who will be using the application, detailing what they want and how they will use it. These are directly correlated into features, assuming they are accepted by the project manager, and can then be handled individually by different people. While very simple, it provides a clear-cut bridge between the users and the developers of a product. So far, it has proved to make things much simpler for everyone and has streamlined the whole process.
In my previous post, I discussed many of the issues and complications that come with working in teams to create an application that may not even fit the end user's needs. It is a huge problem in any environment where the developers are not the people who will be using the application (which is nearly all development environments).
This week they began to roll out a new tool called "Icescrum". It is a form of scrum management that essentially provides a new way to organize the creation of an application.
Essentially, it begins with "user stories" that are requests or "stories" straight from the people who will be using the application, detailing what they want and how they will use it. These are directly correlated into features, assuming they are accepted by the project manager, and can then be handled individually by different people. While very simple, it provides a clear-cut bridge between the users and the developers of a product. So far, it has proved to make things much simpler for everyone and has streamlined the whole process.
Sunday, April 9, 2017
Designing Applications
04/09/17
Big data is a field that is truly exploding in recent years. It has become a bit of a buzzword in the media but it is truly making huge strides in the way we can track and model population-sized data. Dignity Health uses big data to help medical caregivers every day with alert systems, predictive models, and tracking systems for the physicians. It allows them to monitor different parts of the patient's status and alert the physician when they would otherwise not be able to personally monitor them given the large volume of patients in a hospital.
Working with big data applications was something that was a bit of a shock for me; I imagined a timeline for a typical project, even a large one, was on the order of weeks or maybe a few months, but the systems they design can take years to implement. This itself naturally lends to less mobility in larger corporations, as I have observed, but gives me a true appreciation of what goes into these projects that may only have a few core components. When designing a system to handle this much data, they have several different stages and iterations that the product will go through, adding or removing aspects at the request of the end user. It must have very low tolerances for error, especially when this data is in some cases patient's personal data. However, large systems becoming long arduous tasks to create, with a lack of strong direction, constantly changing as there is little way for the end user of the program to essentially draw out the functionality of the application or product. If they were to draw out or prototype what exactly what they want, it would take too long for them or they would essentially create it themselves.
I see improvement possibilities in the way large and small projects are managed in the enterprise. Now, I may not have a full picture but it does seem there is room for improvement in the way our end users communicate what they want out of an application, as opposed to roughly describing it in an email. Dignity Health has been experimenting with prototyping tools and new ways to organize a team under the banner of organizing requests from users so they do not have to revise the product repeatedly because what they initially produce does not exactly match what the user imagined. This should help shorten timelines for projects but also allow for increased mobility as they can spend the time working on new initiatives or actually reworking the applications, rather than remaking them due to not meeting the initial image of the end user.
Big data is a field that is truly exploding in recent years. It has become a bit of a buzzword in the media but it is truly making huge strides in the way we can track and model population-sized data. Dignity Health uses big data to help medical caregivers every day with alert systems, predictive models, and tracking systems for the physicians. It allows them to monitor different parts of the patient's status and alert the physician when they would otherwise not be able to personally monitor them given the large volume of patients in a hospital.
Working with big data applications was something that was a bit of a shock for me; I imagined a timeline for a typical project, even a large one, was on the order of weeks or maybe a few months, but the systems they design can take years to implement. This itself naturally lends to less mobility in larger corporations, as I have observed, but gives me a true appreciation of what goes into these projects that may only have a few core components. When designing a system to handle this much data, they have several different stages and iterations that the product will go through, adding or removing aspects at the request of the end user. It must have very low tolerances for error, especially when this data is in some cases patient's personal data. However, large systems becoming long arduous tasks to create, with a lack of strong direction, constantly changing as there is little way for the end user of the program to essentially draw out the functionality of the application or product. If they were to draw out or prototype what exactly what they want, it would take too long for them or they would essentially create it themselves.
I see improvement possibilities in the way large and small projects are managed in the enterprise. Now, I may not have a full picture but it does seem there is room for improvement in the way our end users communicate what they want out of an application, as opposed to roughly describing it in an email. Dignity Health has been experimenting with prototyping tools and new ways to organize a team under the banner of organizing requests from users so they do not have to revise the product repeatedly because what they initially produce does not exactly match what the user imagined. This should help shorten timelines for projects but also allow for increased mobility as they can spend the time working on new initiatives or actually reworking the applications, rather than remaking them due to not meeting the initial image of the end user.
Subscribe to:
Posts (Atom)