- This topic has 17 replies, 12 voices, and was last updated 1 year, 6 months ago by Tanatorn Tilkanont.
-
AuthorPosts
-
-
2022-09-15 at 9:44 am #38073SaranathKeymaster
Can you give an example of data that you think it could be considered as “Big Data”? What are the characteristics of the data that fit into 5Vs, or 7Vs, or 10Vs of Big data characteristics?
-
2022-09-20 at 8:01 pm #38200Zarni Lynn KyawParticipant
I believe some example of data that could be considered as “Big Data” are
-Patients data
-Data from research studies
-Electronic Health Records
-Data from Wearable devices
-Data from machine in a hospital
-Search engines data (e.g., search query for flu)
-Data from Government agencies
-Payer records10Vs of Big data characteristics
1)Volume – the amount of data collected by the aforementioned sources are voluminous
2)Velocity – the speed at which data is analyzed (e.g., search query for flu) is exceeding the traditional methods
3)Variety – health data consist of both structured and unstructured data
4)Variability – health data is also variable because of the multitude of data dimensions resulting from multiple disparate data types and sources
5)Veracity – unfortunate characteristics of health data, confidence or trust in the data drops if the methods of collection of health data is not verifiable.
6)Validity- health data if collected properly, will have more validity but many data analyst still have to spend a considerable amount of time cleaning the data.
7)Vulnerability – big data brings new security concerns, from what we learn in week 2, data breaches can and will happen. So, providing proper safeguarding measure is essential to promote security.
8)Volatility – in the age of wearable providing exabyte of big data daily, we need to understand the data volatility, so health data can be interpreted in real-time.
9)Visualization – during COVID-19 pandemic, JHU’s COVID-19 dashboard is essential to understand the spread of the disease globally and similar dashboard have been create in local regions to tackle the pandemic.
10)Value – analysis of big health data from economic perspective can lead to best-buy of treatments and in resource intensive settings (e.g., LMIC) investing in best value treatment become a tool to use the resources wisely.-
2022-09-28 at 12:50 am #38382Tanyawat SaisongcrohParticipant
Thank you for sharing the idea. This is very comprehensive example of big data in healthcare.
-
-
2022-09-21 at 5:32 pm #38222ABDILLAH FARKHANParticipant
Big data has already been applied in various sectors, including in the health research sector which explores Genomics data for future therapeutic precision medicine. The growing research of heterogeneous human genomes in the world resulted in abundant information volume retrieved from next-generation sequencing, and that is where big data is necessary. In addition, this kind of data will always be generated along the existence of humans on earth, so the exploration of human genomics-related data will never discontinue.
Like many exemplifications of big data, the genomics data characteristic can reach up to 10 natures such as volume, veracity, value, variety, velocity, visualization, variability, validity, vulnerability, and volatility.
-
2022-09-25 at 2:03 pm #38342Kansiri ApinantanakulParticipant
Thank you for sharing.
The human genomes data is priceless.
Imagine how we could determine whether patient would at risk of develop cancers or NCDs in the future or not. This would allow timely detection and prevention of disease!
-
-
2022-09-21 at 11:34 pm #38229PREUT ASSAWAWORRARITParticipant
I would like to mention about 10Vs of Big Data characteristics:
1. Volume
2. Velocity
3. Variety
4. Variability
5. Veracity
6. Validity
7. Vulnerability
8. Volatility
9. Visualization
10. ValueMy example of Big Data is about information in intensive care unit.
1. “Volume” There is a lot of data generating in the ICU, for example, heart rate, blood pressure, respiratory rate, medication, laboratory data, imaging data, etc.
2. “Velocity” The information in the ICU is generated every second.
3. “Variety” There are both structured and unstructured data generated from the ICU. Structured data includes vital signs, fluid balance, parameters from mechanical ventilation, etc. Unstructured data is those generated from radiology unit, pathological report, etc.
4. “Variability” A lot of data generated in the ICU can come with outliers and anomaly. For example, the blood pressure continuously recorded from arterial line can be abnormal if there is obstruction in the tube connecting between the patient and a transducer.
5. “Veracity” Most of the data generated from ICU has its own reliability. We have to filter bad information out of out basket.
6. “Validity” Most of the data from ICU is accurate and correct, for instance, laboratory data, radiology data.
7. “Vulnerability” The information we gathered from ICU must be kept under highest security due to patient’s privacy.
8. “Volatility” Most of the parameters generated from ICU are relevant. It does not change with time. However, some parameter used in guidelines may be updated following annual guideline review.
9. “Visualization” This is very important issue in presentation information of ICU. How we can made analyzed information interesting.
10. “Value” The most important of such Big Data from ICU is method that we analyze the data and transform it to the valued information.Thank you.
-
2022-09-24 at 4:55 pm #38316Siriphak PongthaiParticipant
Big data includes patient’s data, research study, genomics sequencing data, public records, wearable divide data, search engines, electronic medical records (EMR), smart phones, and social medias.
Characteristics of data, particularly EMR, could fit 10V as following:
– Volume: EMR generated everyday so the quantity of data to be collected and stored are increasing.
– Velocity: since there are hight amount of EMR generated, the time for processing and transferring might be slower due to high volume.
– Variety: most of the time, there are many different types record for example, .pdf file, pictures, videos, or even email.
– Veracity: records must be assured to be used in making decision or intended process.
– Variability: records can be in disparate types, this can cause inconsistence speed of data loading
– Validity: records must be accurate and can be used and analyzed for other purposed.
– Vulnerability: EMR must be significantly considered for confidentiality and security for patient.
– Volatility: the record should be valid because when time passed by, or the next 20 years, the records become essential and needed.
– Visualization: data can be processed and representing in an understandable display.
Data process and act: how data can be representing.
– Value: the data collected can be useful and analyzed for specific outcome. -
2022-09-24 at 8:25 pm #38326Boonyarat KanjanapongpornParticipant
Technology increases ability to collect data so there is much current data which can be considered as big data. Below is my example of EMR characteristic which fit in to 5V’s
Volume: Big data means large amount of data which has been generated continuously. Large amount of data, such as collection of national health EMR, would need effective systems to process, storage and analyze.
Velocity: Data have been increased and generated at a fast rate. This characteristic of big data is useful and would increase the chance to process and make an action in real time. EMR are generated and updated every day.
Variety: Becoming national health EMR, data could have been collected from different sources which have different types of data including structure data, such as syntactic and semantic system for data input, and unstructured data such as random text in medical notes.
Veracity: EMR could be gathered from many resources and many type of data which might be composed of ambiguous and unreliable information. Big data management would help to create the right dataset for quality data-based decisions.
Value: EMR has large amounts of data, however, there are some parts of data which will be useful and could give value or create better decisions in some specific issues. For example, if drug interaction is main query, filtering and analyzing EMR for current medical used and past medical history related to symptom might be priority consideration.-
2022-09-25 at 1:59 pm #38341Kansiri ApinantanakulParticipant
I do agree with you that EMR need the big data management.
EMR is the value source and definitely fell into “big data” criteria.
Proper management of this data would benefits for all healthcare stakeholders: physicians, patient, policy makers and so on.
-
-
2022-09-25 at 1:56 pm #38340Kansiri ApinantanakulParticipant
My example would be data from wearable sensor for example smart watch, real time glycemic level monitoring.
This data fit to “5V” definition of big data because:
1) Volume: The wearable sensor could generate thousands of records of each patient and imagine that the data were pooled. This would create the gigantic dataset
2) Velocity: The record was generated in real-time manner. For example, my smart watch keeps tracking my paces and alert me once I met my daily exercise goal.
3) Veracity: Since the technology is well-developed and sophisticated. The data from real-time sensor is accurate. Sometimes their acceptability may exceed the traditional method of measuring.
4) Variety: The wearable sensor can be designed to collects multiple of health information for example: paces, heart rate and rhythm, oxygen saturation, body temperature or even fall detection in elderly.
5) Value: For me, I think this V is the most important one. The health data is useless if there is no clinical meaning, we could get from them. Nowadays, the data from wearable sensor is acceptable from users (patient) and sometimes acceptable from physician perspective. In the future, we may see the project that develop from real time sensors data. I think this maybe one of the game changers in healthcare technology field.-
2022-09-26 at 11:42 pm #38372Boonyarat KanjanapongpornParticipant
I agree that Value in Big data is the most important V which could generate some changes in healthcare. Acceptable data from personal wearable would generate massive of data and it’s interesting for me as well to see an impact of this tool to health issues in the future.
Thank you.
-
-
2022-09-26 at 1:29 am #38350Hazem AbouelfetouhParticipant
An example of big data is online purchasing orders from websites like Amazon or online markets and collecting customer data and order history to predict products recommendation. characteristics that fit into 5Vs are:
Volume: There is a large amount of data collected from products viewed by the customer and complete/incomplete orders.
Velocity: Data processing, analysis, and recommendation should be fast and in real time.
Value: Customer data stored for a long period to be used in predicting customer behavior and increasing sales.
Variety: Customer data contains many types from categories, geo data for a customer location, product location, price range, etc.
Veracity: Products could be supplied by many companies to customers, and data from these companies should be accurate and reliable.-
2022-10-09 at 4:47 pm #38603Tanatorn TilkanontParticipant
Thank you for sharing. Online marketing is another Big Data that we can see in our daily life and it surely fits the 5Vs characteristics of Big Data.
-
-
2022-09-28 at 1:47 am #38383Tanyawat SaisongcrohParticipant
Beside in healthcare that lots of you already mentioned, I think “GPS and satellite data” used in transportation sector can be considered as big data too. Once a large amount of data (volume) collected through radiofrequency identification sensors, geographic positioning and satellite images (variety) from navigation GPS systems in vehicles and application installed on mobile devices, it provides us with real-time data (velocity) about route traffic and even alters accident-prone areas (value).There are recent technologies that can boost GPS accuracy (veracity).
-
2022-09-28 at 7:39 pm #38397Siriphak PongthaiParticipant
Oh yes! I couldn’t think of that, which I actually use everyday!
Thank you for sharing such a wonderful example.
-
-
2022-09-28 at 2:57 pm #38390SaranathKeymaster
Great examples all! Thanks for sharing.
-
2022-09-28 at 9:36 pm #38399SIPPAPAS WANGSRIParticipant
Can you give an example of data that you think it could be considered as “Big Data”? What are the characteristics of the data that fit into 5Vs, or 7Vs, or 10Vs of Big data characteristics?
We are in the world full of collections of data, which have been stored many years back and is tremendously increasing over time. Since my personal background field was medicine, I can say that healthcare data is uncountable and some of it rarely is being used. For example, summary of non-communicable disease (e.g. Diabetes, hypertension, dyslipidaemia, chronic kidney disease) are being sent to national HDC (Health Data Centre) periodically to visualise and take a proper action. The data format is known as 43-Files, mainly used for statistical and reimbursement purpose. Every hospital is obliged to sent these files for decades. Can you imagine how big these collections are?
These 43-Files collections fit in to these big data characteristics:
1. Volume – this is the best known characteristic of big data for most people. Like I have mentioned, 43-Files collected from hospitals over the years for every visit. You do the math but unfortunately I do not have a precise of how big they are. Let’s say in petabytes, I suppose.
2. Velocity – healthcare data is being generated every second there is a patient encounter.
3. Variety – 43-Files might lack of variety of data because they are stored in SQL (structured data) and exported as CSV files (comma separated values). They contains no images or any binary files other than a plain text.
4. Variability – 43-Files are inconsistent and prone to error due to its nature which comes from various sources, variations in human input format, and a high volume of patient visits.
5. Veracity and 6. Validity –– like I have mentioned in No.4, 43 Files are prone to error, so apparently they appear to have unuseful data.
7. Vulnerability –– 43 Files are unencrypted. Data protection is crucial and sending them back and forth must be done in a secure manner.
8. Volatility — for 43 Files are yet to be determined.
9. Visualisation — Yes
10. Value — 43 Files are intended to use for statistical purpose, but it also contains other data which may be useful upon data cleaning. -
2022-10-09 at 4:44 pm #38602Tanatorn TilkanontParticipant
Talking about Big Data, one example that I could think of is data in Media and Entertainment. In daily activities with numerous digital gadgets, social media platforms generated a large amounts of data, such as in Facebook, Twitter, Netflix, Spotify, Amazon, etc.
In this discussion, I would like to pick up Netflix as an example of the entertainment industry that uses Bid Data analytics, collect data from users, analyze the data and provide customer recommendation.Volume: More and more data is gathered from users watching movies/ series/ TV shows/ documentaries etc. This data gathering can further use to gain important knowledge and understand customer behavior.
Velocity: The data processing is very fast. Netflix will recommend movies/ series after we have finished watching one of them. Moreover, the first time used this application, it asked the information about your interesting types of movies and to pick up an example of a movie that is your favorite. Once done and get into the application, the application will recommend movies that are similar to the movie you like the most. This can be implied that the data is fast processing and maximized efficiency.
Variety: With the variety of video watching from various customers, Netflix can use it to target points of interest and improve customer satisfaction.
Value: As discussed earlier, the collected data could be used to analyze and provide custom recommendations and gain satisfaction.
Veracity: This kind of data is easy to be trusted and can be proved by the customer themselves if it is their favorite. Lots of customers continue to pay for Netflix monthly as it can watch at anywhere and any time and there are many new movies coming every day. People could enjoy a variety of movies and easy to search for their favorite types of movies.
-
-
AuthorPosts
You must be logged in to reply to this topic. Login here