The use of databases is very prevalent in offices where individuals are working on building design. These databases have been used for a long time in accomplishing tasks such as cataloging components, keeping track of customer requests, and creating marketplaces. As big data has become more relevant this list of tasks has continued to increase. Big data describes much larger databases and can fuel much more powerful applications.
The power companies that can manage to collect and store large amounts of data has become relevant through the growth of AI and will continue to become more relevant. The usage of neural nets and other complex modeling methods become more accurate when larger amounts of data are fed into the model. With more instances, the models are able to learn more complex patterns in the data and the impact of extreme outliers is reduced. For design offices, this could mean building models to predict what areas will have the need for new buildings in the future. Additionally, they can predict the future price of specific materials so they can better plan when they should buy materials and what materials they should buy. By accurately making these predictions, companies can safe large amounts of money and make more informed decisions. These examples only scrape the surface of what this type of applications big data can fuel and will continue to fuel. However, storing large amounts of data poses some issues.
As companies desire to store more data, alternative methods of storing this data has become necessary. Traditional ways of storing data have become very expensive and insecure. Storing large amounts of data on a single system poses a serious threat to the integrity of the data if something were to happen to the system. Partially for this reason, distributed storage has evolved. Distributed storage allows users to store their data on shards across a variety of clusters. This avoids the need for massive amounts of storage on a single system and allows companies to store portions of backups across various systems. As these methods become more and more affordable, the storage required to access big data will become more frequent.
Comments
Brian: https://ae-410-510-ay19-20.blogspot.com/2020/02/b5-what-is-sql-and-why-is-it-important.html
I think your analysis of SQL gave a really good overview of what the language is and what it accomplishes. I agree with you that SQL fulfills a very important task in proving a uniform way pull data. Although SQL is primarily designed for relational databases, there are many SQL-like languages that operate very similarly to SQL and work on non-relational databases. Good work!
Amanda: https://ae-410-510-ay19-20.blogspot.com/2020/02/b5-object-oriented-databases_8.html
I liked your thought on OOD and how you were able to make the ideas understandable to those who haven't been exposed to them before. Your analogy with the car gave a very realistic example of how to think of what a class is and how each class has different properties to it. The use of classes within databases and programming also adds another layer of organization to the system.
Mika: https://ae-410-510-ay19-20.blogspot.com/2020/02/b5-mika-awai-sql-in-databases.html
You gave a very thorough explanation of the differences and key uses of SQL and NoSQL. I particularly liked how you discussed the advantages of learning SQL to excel in data science jobs. As someone, who has completed a Data Science internship I agree with you. I used SQL-like technology almost everyday throughout the duration of my coop. I was able to use SQL-like syntax when using pyspark to pull data from NoSQL databases as well. Learning pyspark had a very small learning curve once I knew SQL.
4 comments:
Michael,
I found really interesting the predictions you were able to make in terms of how databases are going to help create more efficient designs by accessing previous data and using AI. I think that this relates to what was mentioned in class for generative design, since AI would be able to access various databases in order to generate a design that will provide the highest performance at the lowest cost.
Michael,
I never thought about buying goods using a neural network, this is a very intuitive idea that could be used in many different industries as well as construction. But then it the question of the accuracy of the neural network because people have tried to predict the stock market but as far as I know, none have succeeded. It is easier to track simple prices over time but, I was just thinking ahead. Also, the majority of the thing you wrote about were similar to ideas I had and I fully support you.
Michael,
It is interesting that you pointed out the use database systems has in storing memory. Like you stated, the bigger companies have large amounts of data that needs to efficiently be stored without the risk of that data becoming corrupt. This way, smaller portions of data can be stored throughout numerous devices that way the memory is safe.
I enjoyed reading your post, it is well structured and makes some good points. I liked that your pointed out how much can be done with “big data” and how hard it is to find reliable space for it. I just wanted to point out another disadvantage of having large databases is navigation. Creating and putting together, cataloging large amounts of information is not an easy task. The way the information is out together directly influence on the way it is accessed. And if it is not done correctly some of the information might be unused.
Post a Comment