63 votes 63 votes An ER model of a database consists of entity types $A$ and $B$. These are connected by a relationship $R$ which does not have its own attribute. Under which one of the following conditions, can the relational table for R be merged with that of A? Relationship $R$ is one-to-many and the participation of $A$ in $R$ is total Relationship $R$ is one-to-many and the participation of $A$ in $R$ is partial Relationship $R$ is many-to-one and the participation of $A$ in $R$ is total Relationship $R$ is many-to-one and the participation of $A$ in $R$ is partial Databases gatecse-2017-set2 databases er-diagram normal + – Madhav asked Feb 14, 2017 • edited Jan 18, 2018 by Puja Mishra Madhav 22.6k views answer comment Share Follow See all 19 Comments See all 19 19 Comments reply Tarun Pandey commented Feb 16, 2017 reply Follow Share According to Madeeasy answer is C and D both. Will this leads to "Marks to ALL"? 1 votes 1 votes Arjun commented Feb 16, 2017 reply Follow Share Even though technically D is possible, C is obviously better. No way they will give marks for both. But I can be wrong.. 12 votes 12 votes Tarun Pandey commented Feb 16, 2017 reply Follow Share Why this question does not leads to "Marks to All", due to ambiguity? https://gateoverflow.in/118597/gate2017-2-45 1 votes 1 votes Arjun commented Feb 17, 2017 reply Follow Share See my explanation now- well there is a slight chance for mark to all, but negligible. Also, mark to all does not give anyone benefit, as individual marks go up, but since this happens for everyone normalized mark goes down. It will only bring those who answered this question as 4.72 less mark :) 1 votes 1 votes Manoj_Kumar commented Feb 23, 2017 reply Follow Share Korth book Page no 289 18 votes 18 votes Ashish Patel commented Feb 25, 2017 reply Follow Share I am still confused between One to Many and Many to One... Any Good Comparisions available? 0 votes 0 votes Ronak Khemchandani commented Feb 26, 2017 reply Follow Share @Arjun Sir Sir i was confused in C and D coz both seemed to be correct but i went with option D, and now the official key says only C. Should i go ahead and challenge because this 1.33 marks could get me within top 100..! 2 votes 2 votes prayas commented Aug 24, 2017 reply Follow Share @Arjun Sir Isn't it possible to capture many to many relationship without a juntion/relation table.Although we will have repeated data this isn't impossible?? 0 votes 0 votes Sayan Bose commented Nov 5, 2017 reply Follow Share Please modify the question. There is a difference between one and own attribute ! 1 votes 1 votes vaibhav101 commented Feb 3, 2018 reply Follow Share This question is gramatically incorrect! The original question said: which does not have its own attribute 0 votes 0 votes CJ147 commented Nov 28, 2018 reply Follow Share D cannot be the answer since there may be a case where a merged table has NULL for pk which is not allowed. 0 votes 0 votes Abhishek Gupta 1 commented Dec 9, 2018 reply Follow Share Why many to 1 not 1 to many?? 0 votes 0 votes ayushsomani commented Dec 23, 2019 reply Follow Share @CJ147 No. It won't have Null Values for Primary Key. Coz we are combining Relational Table on n side. I think C,D are correct but C is obviously much better than D. 1 votes 1 votes JashanArora commented Jan 11, 2020 reply Follow Share My doubt could sound silly as heck, but please clarify. Are we assuming "many-to-one" as A:B is N:1? I mean, it could be B:A as well, eh? Or do we assume A:B is N:1 because of the natural alphabetical order? 4 votes 4 votes goxul commented Jan 31, 2020 reply Follow Share @JashanArora If it's given there is a many to one relation between A and B, it translates as A -----m-----:R:------1------B i.e each entity of A can be associated with at most one entity of B and each entity of B can be associated with zero or more entities of A. 0 votes 0 votes JashanArora commented Jan 31, 2020 reply Follow Share @goxul Well, that's the point. It'd have been nice if the question actually stated what you say it states. 0 votes 0 votes goxul commented Feb 1, 2020 reply Follow Share @JashanArora Yeah, they shouldn't leave it up to interpretation. However, if you see Korth and the accompanying slides, this is the convention that they follow as standard. 0 votes 0 votes Surya_Dev Chaturvedi commented Dec 14, 2020 reply Follow Share As we all confused in C & D. firstly I go through wrong Answer, Which is in our mind. If we take partial participation on A side than also, we have to merge R with A and Only 2 Tables are there i.e. (A,R) and (B). Now Visualize It : If we merge R with A and their is partial participation which means we can put NULL values when there is no Relation from A to B. So lets assume there are lots of entities present in A which does not depends on B, so in relation we have to Put NULL values in place of Foreign key. When We put lots of NULL values and there are only few entities which has participation with B, Than our table become too Lengthy with redundancy , so instead of this we make another table for R because A has only few entities which has some relations with B. So Now there are only 3 Tables (A with foriegn key of R) and (R with Foriegn key of B and Foriegn key of A) and Table B with foriegn key of A) So I think Now you realise why option C is better choise than D. Now you have one doubt remaing. How many minimum tables are required in A has many -to- one relation with B and Both have partial participation ? Ans) 2 (A and R with Foreign key of B) and (B with Foreign key of A). So Doubt is – Here we are taking 2 tables and merging R with A , so why not we merge in this question ? Because In this question it is saying tell me Minimum number of table (so it also include the REDUNDANCY) But when we want our solution redundancy free, than we make different table for R. SO I think it is more clear now. If Yes UPVOTED else DOWNVOTED. 18 votes 18 votes Shiva Sagar Rao commented Jan 19, 2021 reply Follow Share As per GATE 2017 official Key only Option C was correct. If link doesn’t work use archive: https://web.archive.org/web/20210113064019/https://www.gate.iitg.ac.in/2017answers/CS2.pdf 0 votes 0 votes Please log in or register to add a comment.
Best answer 42 votes 42 votes The relation table for R should always be merged with the entity that has total participation and relationship should be many to one. Answer is C. Arnabi answered Feb 14, 2017 • edited Jun 21, 2018 by Milicevic3306 Arnabi comment Share Follow See all 27 Comments See all 27 27 Comments reply Show 24 previous comments ananya_23 commented Sep 30, 2023 reply Follow Share No @khushi, A won't be possible because in 1:M or M:1, Relationship will always merge towards Many side. In option A it is 1:M (1 on A side) so definitely not correct. As far as the confusion goes with C and D, I agree with Arjun Sir's ans above that choosing option with total participation is better than partial one. 1 votes 1 votes KhushiRastogi commented Oct 7, 2023 i reshown by KhushiRastogi Oct 8, 2023 reply Follow Share @ananya_23 , in Option A it is 1:1 0 votes 0 votes ananya_23 commented Oct 7, 2023 reply Follow Share I believe it's 1:M. Please check the options.. 1 votes 1 votes Please log in or register to add a comment.
17 votes 17 votes (C) We need following two conditions: A:B = Many to one A has total participation If A to B is many to one, then for each row in table A, the relationhip can be added as an extra column. Total participation ensures that there is no null values. Kloseup answered Feb 14, 2017 Kloseup comment Share Follow See all 0 reply Please log in or register to add a comment.
9 votes 9 votes |A|===== <R> ---------|B| here relation R between A and B represent many to one with total participation at A entity, this representation is same as definition of week entity, ans is C 2018 answered Feb 14, 2017 2018 comment Share Follow See all 5 Comments See all 5 5 Comments reply Show 2 previous comments PratikDey0316 commented Nov 25, 2020 reply Follow Share Can anyone explain to me how does ‘many to one ’ or ‘one to many’ affect the process of merging the relation R with A? 0 votes 0 votes arjun_harikumar commented Jan 29, 2021 reply Follow Share Consider a schema of EMPLOYEES and DEPARTMENT. The relation is WORKS FOR. Now if start date is an attribute of WORKS FOR relation. Then we associate this attribute to employee table ( n side ) and not to department table, since an employee will have only one start date for a department but department can have multiple start dates for various employees in the department. 2 votes 2 votes PratikDey0316 commented Jan 29, 2021 reply Follow Share thanks for the explanation :) 0 votes 0 votes Please log in or register to add a comment.