wild bear
===

### General Questions ###

investigator: Okay, recording progress. Now you should see general questions regarding usage of digital library. 

wild bear: Yes, I am. 


##### GQ1 #####


investigator: Okay, perfect. So, then we continue with the first question, which tasks do you usually use a digital library for? Please tick all answers, which apply and complete your own tasks, if they're not given here. Please give short or examples of the tasks you are ticking. So for example, you could say, "I tick person search because I like to keep track of myself". So, what your tasks?

wild bear: So, I usually search for papers, of course. So, Yeah, sorry, kinda confused. But yes, I usually search for papers. Sometimes also for persons like in the field of study I'm working right now. And also, I sometimes keep track of myself. Of course, and I tried to get like bibTeX data and in general the citing information from these research portals, whatever. And of course, also I get full papers to read them. And usually, I first scan the abstract if it's actually useful for me. And yeah. 

investigator: Okay. Are there any more tasks you want to include under "Sonstiges", which are not given here, but you also use it digital library for?

wild bear: So, I usually also look at the paper and look if there are similar papers, but it's basically paper search as well, so, maybe not. 


##### GQ2 #####


investigator: Okay. I think then we can continue to the next question. Which system or digital library do you usually use to solve these tasks? Please tick on answers, which apply and name others, which also apply, but are not given here. Please give short oral descriptions of why you like or why you use a specific system. 

wild bear: So, I usually always use Google scholar and I'm not sure if that's the best way to go, but sometimes I also use Scopus or directly science direct search because in my field of studies, they really have lots of papers.

I use research gate to actually get papers, which I otherwise don't have access to and sometimes I also use just Google to search for a paper to get the full text if it's not available directly via the Google Scholar search, for example. And I think that's basically it. 

investigator: Okay. So, no other system or libraries, which are not given here?

wild bear: Yes. True. I'm using connected papers a lot. 

investigator: Oh, okay. 

wild bear: I think I already told you about it so I can put in the title of the paper, and it looks for similar papers, which for example, similar authors, or have a very similar topic with another method approaching the subject, for example. Yeah. 

investigator: Okay, perfect. Anything else or is that it and we should continue to task one?

wild bear: I think we can continue. 


### TASK 1 ###


investigator: Okay. Then you should click to the next page, and you should see task one. 

Consider the following task, find two experts on a topic of your liking. Example topics could be "domain specific query languages" or "hashing functions", but should be from the broader area of computer and information science. The topic you're picking can be anything. It does not have to be one of the two. And you do not really have to do this task right now. So, we are only going to talk about how you would solve this task. 


##### TASK 1.1 #####


So, my first question would be what is your chosen topic? 

wild bear: So, I would choose an actual example. I chose it basically. I was looking at "river discharge by neural networks".


##### TASK 1.2 #####


investigator: Okay. And how familiar are you with this topic? 

wild bear: In general, with river discharge I'm kind familiar, but I didn't look into it neural networks yet. 


##### TASK 1.3 #####


investigator: And how would you define an expert?

wild bear: How would I define an expert? I'm not sure what you are asking. How, how do I define it or how do I... 

investigator: What characteristic has a person who's an expert in your own words?

wild bear: Okay. I would say an expert would have definitely written a paper about this topic, so probably he's working in a research institution, which specifies and in the broader area of this topic and this case hydrology or neural networks. 

investigator: Okay. Anything else you want to add to this definition or is just basically all of it? 

wild bear: I mean, we could use subcategories. For example, like a researcher, like I am, would be, ... How to say it. Like yeah, I'm basically just it's the start of my career. So, I'm maybe not that much of a expert. So, I would probably look for professors and not research assistants, basically.

investigator: Okay. So, if you feel like this really gives your definition, then we should continue to question four?

wild bear: Yeah, I think that's my definition. 


##### TASK 1.4 #####


investigator: Okay, perfect. So then how would you solve the task of finding two experts on a topic of hydrology with neural networks? Was that your topic? 

wild bear: Yes. 

investigator: Okay. 

wild bear: So I would go to Google scholar and for example, put in "river discharge" and "neural networks" or similar search terms and look for papers which have been published in probably the last five years, maybe even the last two years, if there's something available. And then I would also look, as I said, I maybe would put this paper into connected papers and see if there are similar research repeat like every paper, which seems relevant to me. 

investigator: Okay. 

wild bear: So, the task is to find experts, right? So, I would look at the authors and also would basically search the institute they're working in and see if they are, yeah, say a professor, I say head of research team or let's say have lots and lots of literature in this kind of topic. Or if it's like say, just have written one paper and in this subject and yeah, I think that's it. 

investigator: Okay. Where would you search for the authors? Would you search for the authors of the papers in connected papers or somewhere else? 

wild bear: No, exactly. I would search... as I said, my first step would be to identify relevant papers and the second step would be to search for the authors in, and for example, Google Scholar and as well on their institute's webpage probably. And yeah. 

investigator: Okay. And how exactly do you search for their institute? Do you search for the name in like the search engine or is the search institute linked somewhere? And how do you do this? 

wild bear: Usually you just get the name of the author and their university on the paper, right. So, I can put in the name of author and the university in Google, for example, and it will somehow find the correct webpage of the institute.

investigator: Okay. And you mentioned that you would look at the other works of a specific expert candidate, let's say. Would you look at the other papers of the person on their institutional webpage or would you use Google scholar or, yeah?

wild bear: Probably I would use Google Scholar to also search. But probably I would use both, maybe check their research gate as well. That's the institute page, yes. 

investigator: Okay. And then that you have looked at the papers, do you also look at the content of the papers or only at the numbers that are given there? 

wild bear: No, of course I would also, okay. Probably I should mention, but of course I look for relevant papers, which means like I'm looking if they have, say been cited a lot, but especially with papers which have been published in the recent years, maybe I would look at some other indicators which you are way more familiar with. So honestly, I don't know too much about it. Like some... I forgot all the names. But also, I would definitely look where it has been published. So, in which journal it has been published or which conference it was presented at. If it's an open access paper, for example, at arXiv and somehow try to estimate how relevant is paper in my research subject is. 

investigator: Okay. Where do you get the citation information? Is this something you get at the respective digital library that you are at, or would you change the digital library or where do you usually get...?

wild bear: Usually I just use Google Scholar to find out about to the citation information. 

investigator: Okay. And you also said that you would check the venue where it was published. What does the venue give you? Is it like, some type of topical fit or is it like some type of quality or...?

wild bear: Just both basically. I mean conferences always ranked as well. So, if it's a top-notch conference, I know it probably is a good paper, which has been well received. And if it's like lower tier conference might also be not as good. So, I've also been researching topics where there are not that many papers and I also have read like papers where I saw, I'm not sure if that's actually worth publishing and so I always can check if I think the paper looks kind of yeah, not that good. I will always check if it's published in a well-known conference or well-known journal or...

investigator: Okay. How do you know what's a well-known conference? Do you just know in your field, or do you have a webpage where you can check this information or where do you get this information from? 

wild bear: Yeah, honestly, that's one point we already have been talking about. So, Yeah, I just put in the conference and there will be some, I forgot the name of the webpage, but there are some metrics which rank it, but it's always hard to really understand what the ranking actually means. But yeah, you can compare it with other journals or conferences in your subject. And then you can somehow estimate how well known this conference or journal is. Some of the journals you just stumbled across so many times that you just know it's a good journal. 

investigator: Do you want anything else to this question? 

wild bear: No, I think that that's. 


### TASK 2 ###


investigator: Okay, perfect. Then I think we can continue to the next page and the next task it's task two. Consider the following task: find relevant papers from a topic of your liking, which appeared after 2017. Example topics could be "paper recommendation" or "author disambiguation", but should be from the broader area of computer and information science.


##### TASK 2.1 #####


You can, again, pick whatever topic you like, or you can pick the same topic as before. So, question one would be, what is your chosen topic? 

wild bear: So, let's say it was the same topic. So "river discharge and neural networks". 


##### TASK 2.2 #####
##### TASK 2.3 #####


investigator: Okay, nice. So, we can skip question two of your familiarity with it and go directly to question three, which asks you, how would you define relevancy?

wild bear: Yeah, basically the same as before. So, I took the approach to first look for papers and then look for experts who have written these papers, right. So, I would, again, just look for the respective papers and look how many citations they got. What other metrics say about this paper? Maybe even how many downloads they got if it's on research gate or on... I'm not sure. I don't think Google scholar gets it. Also Yeah, I would look at the conference or the journal it was published and yes. I said, maybe look for some other metrics, but usually when reading a paper, also, you get kind of a feeling how good this paper actually is. 

investigator: Okay. So, you find relativity of a paper as what?

wild bear: Okay, I'm sorry. So, relevancy basically is how well received it is in the scientific community and of course, how well it fits to my topic of interest. How well it is written properly. And so, yes. 

investigator: Okay, nice. So, if this is your whole definition of relevancy, I think we can continue to question four, yes? 

wild bear: Yes. 


##### TASK 2.4 #####


investigator: Okay. How would you solve the task with your chosen topic? How would you find relevant papers on the topic of river discharge and neural networks which appeared after 2017? I think you already mentioned part of it, but yeah, let's just start and you tell me how you would do it. 

wild bear: Yeah. Again, I would probably use Google Scholar. Just put in the search terms like "river discharge and neural network" maybe use parentheses. And yeah, of course I would then check in from the last five years. So, it's from 2017 on and yeah, then I would see what papers come up. And if I feel like they are not that good papers in it, I would probably vary my search terms a little bit. For example, I would use "runoff and neural networks" or maybe be more specific as well. Maybe I put like "convolution neural networks", for example, if I want to go into that direction. And maybe I would also use Scopus, as I said, if there are not as many papers appearing as would have hoped to, and if I found good papers, I would also use connected papers, as I said. I really like this tool. And then I would be able to see which other relevant paper are there in similar topics.

investigator: What would be an indicator of your resulting papers being not so good papers? 

wild bear: An indicator for it. You probably... I always quickly read the abstract and if it is unclear what the paper's about after the abstract, I always think like sounds kind of fishy or I'm not as deep into the topic as I should be, could be both of it. And then I would probably, again, look at where it has been published, how many citations it got. Yeah, if it was presented at a conference and at which conference look for metrics and maybe, probably read the method part and if it's not well written, I would maybe look for another paper and maybe come back to it later, but yeah.

investigator: Okay. Where would you read the abstract? Would you read it in connected papers or in Google scholar? Or would you download the paper, or would you search for abstract somewhere else? 

wild bear: Usually I just click the link on Google Scholar, assess the abstract. And if it's interesting, I download the paper. 

investigator: Okay. So, you download it from the linked page and this is like another digital library or is this still Google scholar?

wild bear: No, it's usually somewhere externally, but it differs a lot. It might be research gate. It might be the journal itself. It might be wherever. 

investigator: Okay. Nice. Okay. And indicators for good papers are then the opposite of not so good papers and I guess?

wild bear: Yes, of course. Yeah. 

investigator: Okay. So, you would first check all your indicators, like the abstract, where it was published, how many citations, assess if the methodology part seems okay-ish and other metrics, whatever, before you decide if your query resulted in good papers or bad papers?

wild bear: I'm sorry before, well usually I just check, if I got relevant titles. So, if there's lots of titles, which seem relevant to me, I go into the next step to explore the papers. 

investigator: Ah okay. I misunderstood. Okay. So, you do your search with your terms, then you check the titles. If they seem to fit your topic?

wild bear: Yes. 

investigator: And this makes you then either reformulate a query or going deeper into the papers and checking if they are relevant with the afore-mentioned abstract checking, published where, how many citations and so on?

wild bear: Exactly. So, the first step is of course to check the titles, if they are, seem to be relevant. Yes. 

investigator: Okay. Yeah, I don't know. Is there anything else you want to add to this task, or do you feel like this process, what you described depicts your process fully? 

wild bear: I mean, it always, it's a little bit different when actually doing the tasks than describing it, but I think that's it, yes.


### Thank you ###


investigator: Okay, perfect. And then I can stop the recording.

