Mastercard Interview Question | Map-Reduce Concept
Question
Related questions
The Given Data set, Table has transitive dependencies. Explain Which Attributes are partially dependent. How will you overcome this? Explain the approach?
Write a query to Plot a Relationship between the highest followers in every country (Country Wise Highest Followers) and show which country has the highest followers and which country has the lowest one. Explain the approach?
Write a Query to Find the account with the highest number of followers in every country. Explain the approach?
Assign Rank To each Category by authentic engagement in each country and return the category with a rank less than 3. Explain the Approach?
Write a Query to check for Missing values in the given table? Explain how will you deal with the missing data for qualitative and quantitative data?
Answer ( 1 )
Map reduce is a framework that was developed to process massive amounts of data efficiently. For example, if we have 1 million records in a dataset, and it is stored in a relational representation – it is very expensive to derive values and perform any sort of transformations on these.
For Example In SQL, Given the Date of Birth, to find out How many people are of age > 30 for a million records would take a while, and this would only increase in the order of magnitude when the complexity of the query increases. Map Reduce provides a cluster-based implementation where data is processed in a distributed manner
The biggest disadvantage of MapReduce is high latency. which makes it unusable for real time applications. There are frameworks on top of hadoop like HBase that are well suited for real time use cases too but plain hadoop/mapreduce doesn’t provide that functionality.