Course talk:CPSC522/Identity Uncertainty

Hi Bahare

Nice to read your wiki page regarding identity uncertainty and see your wonderful remark and comparison about these two papers which not only focus on records but also on attributes. I am very interested in the household example in your wiki page because I think this topic can be widely adopted into real world application. I am wondering whether you can add some real world example regarding the identity uncertainty. I am sure that that would be much better for better understanding its real usage. Thanks

Regards Arthur

Hi Arthur,

Thanks for your feedback.

The problems of citation matching, person identity uncertainty, and the explanation about merging companies are real world examples. Did you think of something else for real world examples?

Cheers,

Bahare

Hi Arthur,

One other thing is that both the papers worked on real world examples. The whole field is about real world examples. There is one example of merging companies which I'm working on a similar one in a project.

I will be happy to hear from you!

Hi Bahare,

Thanks for your kindly update, I think that will be enough for now. Thanks

Regards Arthur

Hi Bahare

This page is easy to understand, and the introduction part is great. I have several suggestions:

1. The reference is displayed not correctly. I think it is better to update the reference mark.

2. The first example is not very clear to me, especially the Figure 1, I believe it would be better to explain what's the mean of D11 D21 or A11, and why they don't just use D1 D2 or A1, maybe using a real example that would be better.

Sincerely,

Junyuan Zheng

Hi Junyuan,

Thanks for your feedback.

1. I just got how we should reference. Thanks! I will change it.

2. By first example do you mean the first paper? I will explain more for sure.

Cheers,

Bahare

Hi again,

1. I update my reference list. 2. I add explanation about your confusing.

Thanks!

Hi Bahareh,

First of all, let me apologize. I totally mixed up the due date of critiquing with the due date of Final Draft. Let me thank you for contributing in our Wiki pages, and what a fantastic job you have done. I really liked all the aspects of the page, its sufficient and minimal. It is also well-explained. Only, I feel like adding some hyperlinks for some of the less intuitive concepts will be great, as the readers can understand the whole thing easier. Other than this minor pointer, this is a great page in my opinion.

Cheers,

Mehrdad Ghomi

Hi Mehrdad,

Thanks!

I will put more hyperlinks before finalizing my wiki page.

Cheers,

Bahare

Hi Bahare,

Very nice job. I also read the two papers, and I think you really made the two clear and easy to understand.

I like the two papers that you used, one focus on records, another focus on attributes, and this is also what I am thinking. Based on different conditions, as you mentioned in your page, different solutions should be adopted. For example, for records with same data structure, we can consider to deal it from attributes, while if not, we must use other ways, like trail matching or like iterative or sequential ways to predict the relationships between records.

How do you think the difference between record linkage and identity uncertainty? Do you think actually they are the same task or there is a little slice different? I think you must have seen “record linkage” many times when you are doing this.

Hi Dandan,

Thanks for reading. It was my first try to go through the identity uncertainty field. It seems like the identity uncertainty and record linkage are the same. The only different that might be is that: in record linkage some knowledge of databases may be used, which I have not seen any in the two papers that I wrote. How do you think?

Cheers,

Bahare

I think the the identity uncertainty is like the core problem in record linkage, as you said, it does not only need to solve the problem of uncertainty, but also need to consider about others like efficiency accuracy and also error handling.