diagrams II - Githubissues

phf1001 / semisupervised-learning-in-cibersecurity

Automatic phishing detection using semi-supervised learning algorithms. Check out web analyzer!

https://krini.herokuapp.com/index

GNU General Public License v3.0

4 stars 1 forks source link

Closed phf1001 closed 1 year ago

phf1001 commented 1 year ago

@alvarag Hello!

I had a meeting with Jesús Maudes and we found few issues on the diagrams.

[ ] One atribute could be moved on the ER
[ ] There is a ternary relationship that I had not detected
[ ] The 'loops' aren't real 'loops' since the admin is making really different roles on each relation, so they are OK!

I have corrected those mistakes and I am attaching the new diagrams:

2 er_krini_v-2

2 krini_relational_v2-0

Also I am attaching the data dictionary:

[ ] I asked Jesús if my dictionary is OK, and he told me that I could have used an ISO, but even though is more than enough. What do you think? If its better to use an ISO, may I have the link to the template?

I would be really grateful to hear your feedback about everything.

alvarag commented 1 year ago

I have to checked carefully all the documentation that you have attached but it looks nice.

In my humble opinion the change to ISO is not essential and does not give you enough to justify the change

alvarag commented 1 year ago

I have been thinking about your questions:

The ternary would not be necessary because of the explanation that you gave to me the other day. In fact I ask you about the need of keeping a history of the instances and I think it would be simplified. Anyway is not bad at all.
About the data dictionary, maybe is OK as you have it but I think is preferable to have a column "key" in which you add whether an attribute is a FK, PK...
You can make the tables \small for making a bit smaller

phf1001 commented 1 year ago

@alvarag Hello!

The ternary is to keep a history of the instances "reported". It is basically the same as the other day, but the other day I was using an array instead (so I used to append dates to the end). The behavior is the same as the other day but I think using a ternary is the most "clean" way of implement it. What I don't want to keep a history of is the "revision" (admin) dates. Anyways, if I don't use the ternary, that means that no user can report the same URL twice (or the report will be invalid / delete previous reports).

Maybe is better to talk about it on wednesday 😆

Data dictionary / \small: OK! Will fix it on the next sprint (after we "agree" on the whole data model :P)

Thanks!

alvarag commented 1 year ago

All right, we'll talk it next week!