Transcript: The Landscape of Academic Literature in Quantum Technologies

hi everyone uh welcome to the 10th uh session of uh quantum conversations uh i'm mart i've been managing uh quantum conversations for a while now i graduated from stanford in march and i've been a member of the stanford quantum computing association i've been running the partnerships there for a while but now josh is doing that i don't know if yorkshire's around right now hey josh do you want to uh introduce yourself yeah um hi everyone i am a uh junior um any leader so i guess about to be a senior at stanford um yeah right now um i run the the partnerships and events committee uh took that over for mert after he graduated um yeah and we've been working together on quantum conversations for uh this quarter essentially and a little bit of last one yeah cool thanks josh and we have alexi who started quantum conversations yeah do you want to do a little intro alexa yes thank you mer thank you very much for for running this uh so i'm alexi kravrov uh i am now the technic like a system development lead at ibm quantum so i'm actually kind of got inside uh after starting this one of the conversations as a community organizer i'm a long-term ibm community partner and my goal now is to basically foster like a system around quantum we have this uh hubs model where we have different universities government labs so we have urban we have uh multiple universal state in the u.s uh university of chicago we have um members of sherbrooke in quebec uh uh in germany we have uh uh iberian non-attacking in braga portugal have uh kasich in madrid from hopper in germany which we just you know launched a new computer we have winners of tokyo and so the question is really for me right beyond the machine machine ibm has the best quantum machines and the first available or the cloud but what's like a system what are the people doing with them you know what are they doing to achieve how can we help create the community so so it's kind of really great to to steal here uh because you know community i think is driving all of this so you know uh thanks for joining and thanks to the speakers thanks for the american euros organizers looking forward to to learn more thanks alexi uh i mean maybe i should introduce quantum conversations a bit too so we gather uh on different topics ranging ranging from um academia which is what we're gonna do today and then uh for example we had industry uh last month uh we talked about quantum finance and i wanna just briefly mention that quantum computing penetrates every area uh that that technology is in right now that ranges from academic academia to industry so research labs uh startups big corporations so previous after i finished my master's actually during my master's i was working at a startup uh that was focused on quantum computing now i'm working for a large finance company and i'm part of their quantum team so we see quantum computing appearing everywhere across the board and that's why we try to keep these conversations very very uh diverse and today we have uh ziki who's going to talk about the academic landscape of quantum computing so zeki do you want to introduce yourself thank you man and thank you for inviting me to this uh series of conversations so i'm zeke zesker i'm original from turkey i'm doing my phd in physics here but also i'm interested in the science and technology policy studies of quantum technologies and today i will talk about the a bibliometric study that we conducted regarding the academic landscape of quantum technologies uh should i share my screen uh yeah just a second so would you like to know our other participants and yeah of course if you can like introduce yourselves just maybe tell us where you're joining from and why you're interested in quantum computing or quantum conversations our session uh that would be great so can we start with you i think that's frozen um let's start with hello hello yeah yeah sorry you were frozen go ahead yeah yeah hi everyone uh this is kiran and i've been uh part of quantum conversation uh probably by thanks to mert and alexey um it's been a platform where i could learn more and currently actually i'm working in an institute in an educational institute [Music] um i think we're losing you a bit or is it just me yeah we are losing uh learning deep learning and computer vision but uh hello yes sorry it's it's like the connection is lagging a bit uh i heard deep learning and uh yeah yeah okay am i out of it now yeah it's better now maybe let's try again and if you can like okay yeah i i i'll just be brief uh like i've been partnered i'm currently working in an institute like in india educational institute university in the domain of deep learning and computer vision but uh i haven't been interested in uh quantum computing uh i've been doing like some projects of my um from my site remotely so probably quantum conversations has been a platform where i could get more about the domains and what the current trends you know so looking forward thank you yeah okay great thank you uh for the introduction we have a very uh we have people all around the world here today huh uh yeah i'm joining from new york zig he's joining from ankara we have two bay area participants actually are you in the area yeah okay cool abyss car can we get to know you yeah hello everyone i'm avishkar i'm joining from nepal uh neighbor country of india okay yeah i am a part of uh one quantum nepal uh it's basically a chapter of one quantum uh one quantum is the google organization and it try to like unite or unite quantum leaders all together and do some uh tangible works in field of quantum company and in revolution to the phantom bombing so i am a project head at uh one quantum chapter in nepal and i'm joining now from nepal and i am high school graduate just graduated 20 in 2020 from high school and i'm so much interested in quantum computing that's why i'm following uh quantum community group group of stanford university since one year great great thank you um yeah it's better to get in in this space earlier the earlier the better um yeah we're very familiar with one quantum actually i've been collaborating with denise and andre for a while now i was like also that uh the one quantum africa session so there was there's a lot of uh a lot of one quantum uh subgroups now uh okay you were in like one quantum africa like meet up i think uh one week ago were you there sorry yeah no i wasn't in the in the meeting i was at an interview with farah they like the series [Music] thanks uh so we have saliha salia you want to introduce yourself hey uh i'm i was just curious about the abstract i'm joining from istanbul okay nice to meet you all nice to meet you nice to meet you and nathan hey nice to see you again uh likewise can you hear me yes we can hear you oh great great yeah um alive well double vaccinated uh survived covet obviously and uh which was a big deal for me giving my age a medical condition right ah so uh triply glad to be here great uh by the way are we doing that lightning session at the end yes yes he's gonna give a speech uh about his work now okay i did uh something recently uh that i mentioned in the uh registration form i never got a response i wonder if i can get two three minutes to show it off too time for me of course of course yeah no worries it's cool great okay okay um all right i think we can get started do you wanna head the way yeah sure uh just let me share the screen and it's going to fall okay can you see the presentation okay thanks so today i will talk about this uh it is actually in early access for the international journal of country information and it has been on archive for the last i know one and a half years so you can easily access uh on the archive version the journal version is an early access if you can access the journal you can check it from there as well it's the same title the landscape of academic literature in quantum technologies uh so i'm from the middle east technical university physics department i have been working there uh for the last i know six nearly six or seven years also i have this second master's degree in science and technology policy department which is the i think it's the only uh policy department in turkey that is focusing on technology policy and science policy and also i'm one of the co-founders of turkey which is part of this uh larger network called q world and i will also talk about keyword very briefly in the end of this presentation so the outline of the presentation is following i will give you just a very brief background i will introduce the methodology and what are the bibliometric tools uh how we collected the data set the keyword analysis the burst detection uh the core corpus and i will present you just a brief discussion uh and the background is actually from my co-authors arsenal i don't know who is a pro professor at the science and technology policy department and he did such a work for nasa astro biology institute just a couple of years ago and while i was taking uh the bibliometrics course from him he suggested that hey not why not we are doing this for uh quantum technologies as well so the original title was a quantum information technologies but later we changed just quantum technologies and so i think most of the people here know what web of science is or you know how citation networks work but to be complete uh the idea of bibliometrics is not a new one it has been around for sorry i don't know what the did you say deborah science web of science web of science so i i i don't know what it is okay please if you have like questions that we have like a discussion so yeah if you have questions please questions like i do yeah so the concept of bibliometrics is something like this it has been around for about 50 years the originator is someone called eugene garfield and the original idea was that to create a sort of a metric that will allow libraries to decide you know which journals to buy and which journals to not buy because by the mid 60s there was this uh let's say uh explosion in science that new journals were coming up and the especially the university libraries had to decide between different journals whether they are being useful for their uh employees or not or the academics or not so using graphic developed this method that if you know let's say you publish an article and it has been cited by certain other articles in certain journals and none of your academics read those articles and they are not using those articles so you shouldn't you know buy those journals because that would be expensive to buy the average by every journal and this actually uh transformed into what we know as you know sukopus web of science uh google scholar is kind of like that but it is much less uh it is of much less quality and all the things that we hear like the age index impact factors the quality factors all come from this area of scientometrics or bibliometrics because it has suppound from the bibliometer studies and for our methodology for this paper we had our initial query which is just something we created we analyzed it uh we signed it to some experts in the field we got some feedback we devised this new query analyzed send it for final opinion and then finalize the study and also on web of science there is this uh let's say option to use highly cited articles which are articles that are not just not articles that have the highest number of uh citations but the that has been getting those number of citations in a very brief time so they are considered as hot particles in that sense and we use two software tools one of them is called closed fever i will introduce it at the end and the other one is also size square tools and i will also show you the website of that tool so the initial query was like this so we got you know quantum inspiration imaging sensing anything that distinctly identifies as a part of quantum technologies literature because there is this difficulty when you are using keywords if you just type quantum you will get the entire literature that contains the word quantum in either the abstract or title or keywords which is something you don't want because you know quantum is used in other contexts as well but if you just type quantum physics or quantum mechanics the articles that you will get are they can be unrelated to quantum technologies as well but also they can maybe not include something that is focusing on quantum sensing or quantum circuits or content cryptography and we actually use this initial query and for the final query after the expect opinion it has expanded a little bit you can see it from here also you can check these keywords in the article as well all of them are included in the appendix so initially we run this analysis for uh titles and abstracts and keywords for articles and proceedings we got 4 40 and 41 000 articles and proceedings and we run an initial analysis we send it to the experts and we used five expert opinion from our colleagues and then uh with their feedback we used the final query and with the final query we only included articles we didn't include proceedings and we run the announcement 6th of june i'm giving you these dates because for example if you use the same query right now you will get more articles but if you limit the date to 6th of june you will get the same data set that we had so these kind of words are actually repeatable so everyone can run their own analysis and for finally we actually excluded articles that had either zero citations or zero references so if an article was not referencing any other article we didn't include them and if an article had zero citations we also didn't include them uh we checked uh whether these zero citation articles were new because yeah i mean if an article is new it will have very naturally zero citations but most of them were uh older articles with zero citations so we thought that you know either they are you know kind of irrelevant to the literature or we can basically exclude them citations do you mean they were cited by other papers yes exactly exactly yeah so let's say let's say you have a paper and no other papers cite your paper and it has been around for 10 years uh then i mean i don't know how fair is this but we thought that they are outliers or they can be excluded from analysis in that sense we can discuss whether this is an uh reasonable option or not and we found that there had been uh journals with five or more articles from the data set there were 495 journals in the data sets so this is the number of articles as you can see there has been a clear increase in the number of articles and this has been a rapid increase uh since the early 2000s and i mean after source algorithm but especially after early 2000s the number of articles has been increasing and also the number of authors per paper have been increasing uh so even if you exclude like the major papers with lots of let's say authors even in that case the number of orders uh per paper has been increasing steadily and the mod of number of orders per paper has increased to three since uh 2013 which means that the huge chunk of papers here uh contain three alters so the field has been becoming more collaborative uh for the say last decade or it has been becoming more collaborative since its inception but this is a general trend in most of the papers and most of the fields but in quantum technologies the trend has been actually stronger in that sense so this is the initial keyword map that we created from our first uh let's say query and 41 000 articles and we actually sent this to the experts to get some feedback you know whether this looks reasonable do you see something that is not supposed to be here and whether the clustering of the algorithm uh is actually suitable or appropriate and i mean it looks more or less uh reasonable because most of the red cluster here actually focuses on the physical hardware and experimental aspects of these systems the green and the yellow clusters are more focused on quantum computing and theoretical aspects the purple cluster here is kind of more focused on quantum communication and the blue cluster here is quantum cryptography and usually quantum sensing or single photon based systems and after we switch to the final cluster we actually limited uh the keyword clustering to a 60 version which means that we only took the top 60 percent of the keywords so to get this more clear clustering in the literature and again you can see that the physical uh implementations of quantum technologies formic clusters uh the quantum communication quantum cryptography and sensing fields form a second cluster and quantum computing and quantum formation science forms a third cluster and if we raise the keywords to the full version 100 version you can see that everything becomes uh closer to each other because we include all the keywords but also effort clusters is can be distinguished here so the quantum computing cluster and quantum let's say more quantum information theoretical clusters gets distinguished which is something expected because there has been this for example quantum discord and entangle entanglement dynamic and quantum phase transition are here which can also be taught as more information theoret or quantum information theoretical concepts that are not related to quantum computing like quantum thermodynamics is an emerging field so it is not exactly quantum computing so you can see that it is emerging but again from here it is more closer to the computing encounter information science than to physical implementations um as i said you can see that there are three clusters from communication cryptography cluster quantum computation information theory cluster and physical realizations of quantum systems and as you increase precision uh these clusters become more compartmentalized and you can distinguish you know more precisely and the second analysis we run call is called burst detection uh which is an algorithm that has been developed like two decades ago and it identifies sudden increases in the frequency of works uh it is not a simple frequency analysis you can check the algorithm that's why i'm giving it here also you can find it in the references you we use size square tools to generate these bursts or to identify these first and in the software that is this density parameters if you decrease the parameter you will only be able to identify uh let's say stronger bursts if you increase your density parameter you can get a smaller burst in that sense so this is for the lowest uh that's the parameter analysis we run from here you can see certain uh fields emerging and then either becoming mainstream or fading away you cannot just by by looking at these you cannot identify whether the field has saturated or faded away but i will give you another example in just a moment so here you can see in uh quantum computation obviously started bursting after short's algorithm and then you know teleportation entanglements quantum computer quantum and following 2000 so i i think someone is speaking uh yeah thank you and here you can see for 2008 there is this huge burst of quantum entanglements you know some more theoretical return quantum optics then quantum discord and finally quantum correlations uh have been bursting in that sense and for the higher density parameters uh you can see the full list in the both in the archive version and the paper online but here are some that have been still bursting at the time of our data analysis so quantum coherence simulation image processing networks annealing steering one of the things that surprised me was ads cft correspondence here but when i spoke with my friends that are working in string theory they basically said that oh adcf is everywhere now and the basic methods for let's say the basic tool of abs cft is dimensional analysis you can play with dimensions you know increase them to freeze them and also that has been useful for quantum computing especially in the terms of control correction and how you work if you are trying to let's say an information theoretical question on black holes uh then you can combine these fields of quantum information science and uh these tools that are using abs safety but as i said i was not able to let's say identify whether a field stagnated you know became mainstream or a keyboard or faded away just by looking at the burst analysis so i run my own frequency analysis and i choose these i choose to compare these keywords here you can see you know any link variance simulation from the list here you can see that they have been bursting and also from here you can see they have been bursting for quite some time so the frequency of these keywords appearing in the literature uh have been increasing and again here you can see that these have been increasing but you know you can you can see quantum teleportation is here and it has not been bursting and here you can see why so the quantum teleportation has become mainstream it it has been steadily like around one percent of the literature or let's say 0.008 percent of the literature so it is there people are using this uh these let's say keywords in their papers but it has been used at the same amount uh almost every year so it is not bursting and finally you can see the the keyword of quantum computing here and again at 2008 2009 there is this huge spike for quantum computing i have been discussing a possible reason for this for quite some time uh there are many possible explanations of this some of them some people say that it's related to funding some people say that at that point of time there was this economic crash and after the crash people just tried to publish what they got some people say that oh it is related to you know super conducting qubits or i mean you can probably find a lot of explanations for that so for the highly cited articles uh for on web of science uh we we use the same query and found published articles and out of these there were 34 articles which had over 30 citations in the set of highly cited articles uh to compare we looked at their respective citations in the full set of are 42 thousand articles and here you can see uh the full list also again is given in the paper so this is the compared most highly cited articles in those sets so this doesn't mean that nielsen and trunk has 6929 citations this means that out of these 42 000 papers almost 7 000 of them are citing minnesota and chunk and out of these 808 papers uh 113 of them are cited miss mentioned so it is the most cited uh academic work in both of these sets but the rest of the set is different for example okay no worries it is it's not the textbook right you're referring to the page it's the textbook it's it's the text people are typing the textbook yeah okay okay yeah uh and you can see that the most of the uh papers for the highest cited papers in the curator set uh are like spanish eckhart gg and epr papers but for the highly sighted set it is kimball walraft niagara and broad hq so it's the quantum internet paper of kimball it's the super conducting qubit paperwork valera i think this is the topological cubits paper by neck and it is the if i'm not mistaken this is uh quantum measures and quantum entanglement paper by horizons uh so they differ you know for the highly cited works are citing different fields than the field itself the field itself is mostly citing the cornerstone papers but the highly cited papers are citing more recent and more uh transformative i mean the papers that are transformative but they are more recent not like the original epr paper uh yeah i don't know why this is here sorry and finally we uh not finally but as at the third level of analysis we looked at the country levels and here you can see that china has the most papers this is followed by the us and if you combine europe like germany and france italy spain even if you exclude the uk the number of papers are almost the same so it's like three times it's like uh 10 000 or on the on that order like 13 000 papers so there is no let's say a clear winner in terms of the quantity of papers and when we looked at the co-authorship analysis of countries with more than 10 published articles you can find this collaboration network and also the software clusters these countries as well so here you can actually see the if let's say you publish the paper and the authors has a affiliation in china and you have an affiliation in italy then there is this uh a single link between you and as the number of authors that is the number of papers that have with authors with affiliations in china and italy this link is becoming uh stronger and the sizes of these nodes are basically the number of papers published in those countries uh so in this analysis singapore is like you know pay effort uh so i can i couldn't uh introduce it to here but uh again you can see the clustering like this is these are the countries that are uh collaborating with china these are the countries like collaborating between germany and france russia spain uh uk so you can check the full list uh from the paper and the final uh analysis was the institutional level analysis so these are the top 12 i think so i think chinese academy of sciences is kind of an outlier here because at least half of the authors from china have multiple affiliations so it's not like 3 300 academics are working actively for the chinese academy of sciences that are working on this field but let's say used uctc has some academics that are that have affiliations in chinese academic sciences as well against the student numbers in the number of students in the no no these are the just just the numbers of articles okay because i mean like if an institution has more researches it's like then it might publish more articles right so um yeah sure sure but this is not a ratio this is just pure numbers uh but even even in the case of pure numbers it is not like uh chinese academic sciences have this many uh dedicated researchers but most of the people who are working in projects at china have also chinese or chinese academic sciences affiliations that's uh something uh a similar thing can be said for i know max planck as well maybe i'm not that familiar like i think a very good like this came out just by the numbers and i mean just by like knowing about the universities we know like some of the top universities in quantum computing and quantum information sciences are included in this list so very interesting thanks and we again run this key original um clustering analysis uh for institutions and if an institution had more than 150 published articles in the set uh so again you should always note that we are running our analysis only on our curated set so it doesn't mean that these institutions have many papers or many restructures but it is focused on this academic literature in quantum technologies and here you can see that the chinese universities and or institutions are clustered against these other groups and also i think this is taiwan and again you can see here there are different clusters uh like japan japanese institutions and mostly the european institutions there is this canadian and australian institutions and there's this american institutions uh but i think one of the reasons the chinese institutions are uh let's say more uh separate from the rest of the world is the double affiliation issue that i have been talking so let's say mart has been working in uh shanghai university but also he has an affirmation at the i don't know chinese academy of sciences so the software considers this as a collaboration because there are multiple affiliations in the paper so it it distinguishes you from these other institutions but even if that is accounted still there is clear indication that the chinese institutions are more uh cultivating nationally than internationally so the discussion is i mean you can use bibliometric tools to examine the field and it yields some variable insights the bursting topics evolve away from straightforward quantum computation into more diverse areas especially like quantum thermodynamics quantum correlations uh and i mean more theoretical concepts like ada cfd or i don't know whether it is theoretical or not in that sense in that particular sense uh china u.s and eu even excluding uk have similar sizes of contribution uh to the academic literature this so this is only uh with respect to quantity uh we cannot i mean we can but we didn't use a qualitative uh way to distinguish these contributions but quantitatively you can say that these three regions of the world are producing almost the same amount of papers or produced same amount of papers until the moment we collected our data and chinese institutions tend to collaborate with each other more heavily rather than internationally and you can see the references here you can check our paper also you can check linebacker's paper for the algorithm that we used and i want to thank you for your attention but also now i will show you very briefly so this is the boss reverse tool that we use i mean you are probably familiar with these uh let's say interesting visualizations you can actually use lost viewers it's a free and open source software uh you can create this by yourself it is very let's say so it is easy to use it is kind of memory intensive so you need some ram in your computer to actually analyze huge sets so for example in my laptop we couldn't go further than 60 thousand art articles uh because it was constantly giving that you know out of memory errors but you can use some tricks and use velcro and also for sci square tools you can use it and you you can do this burst analysis you can do other kind of analysis country level analysis uh it is again an open source and free to use software and the manuals for these are very clear and you can just look at you know you can check the github page as well and finally this is keywords we are part of this q world organization it is it is an international organization with many members around the world so here you can so here you can see our number organizations we mainly focus on education we are organizing these q bronze workshops and here you can see a list of these events and also we is in just a week we will have this keyword quantum science days you can you can check the website so we have these events coming up we have these thoughts we have this q intern project we have these training workshops uh it is entirely free everything is for participants and also if you are interested you can just send us an email or look at the website and we would be happy to topic you and i want to thank you for your attention and thank you perfect thank you zeki i saw alba is one of the speakers in the in the the q world quantum science day yeah yeah she was also a part a speaker at quantum conversations a few months ago oh that's great um okay so uh yeah we are opening up the questions now um i i want to mention some something so actually it's tied to a question but abiskar i think you don't have to raise your hand just shout out to your questions so you go first yeah okay thanks uh okay i have a short question for jackie like can you show that presentation uh that cited list enlisted listed from uh like institution cited i think sorry which one institution cited this one right yeah okay okay uh that is institutional labels analysis and you mean the the the listed institution have many many many researchers or uh many articles are cited by a single researcher uh so this is not for citation this is for authorship so let's say you publish a paper and you wrote one quantum as your institution if there are let's say 1 000 papers that use one quantum as the affliction it would have been on this list yeah so it is not just a single researcher you know and it is not a citation so here uh again we used for authorship you can you can actually do this for uh co-citation as well you know which uh institutions are citing each other but as i said uh such an analysis requires uh a lot of uh let's say computer power in that sense or you can use our or different kind of analysis but we were not interested in that but you can actually do do such analysis as well you can see which institutions are cited the most rather than which institutions are creating the most it is it is entirely doable you can you can do that okay sure thanks thanks okay thank you thank you yeah i i want to ask about your uh assumption of like um not not uh including zero sighted papers so papers that i haven't been sized before and i have an example so low grover uh the the founder of the famous grow research algorithms um paper in in the thesis paper he wrote wasn't it wasn't cited many times i think it was cites two times or maybe zero times i'm not sure so like just as an encounter example and for fun like what was your what was your assumption when you just like disregarded all of those papers have you looked at like trying to include them were they like saturating the data and making your data like not not meaningful so the reason we actually did that was we were trying to find the most productive institutions and how the connections between institutions are formed but if let's say a paper has zero citations it can be two things either it is really new uh or it is it has not created such an impact uh in the literature so we decided to leave them out because if they are really new let's say we run the same analysis in two years they will show up so it's not a big loss and if they are just you know in an academic sense junk then we shouldn't include them but if let's say so there's this new way of analyzing literature is emerging it's called altmetrics so even if you have zeros zero academic citations maybe people are activating your paper maybe people are using your paper in presentations maybe people are using your paper for open source projects they so the bibliometrics or the scientometrics community is trying to incorporate that uh into their analysis as well but the automatics tools are not that well developed to be used for let's say an analysis like this so we decided to leave them out and when i checked the papers i i checked like 1 000 of them very quickly you know just to make sure that we are not as you said leaving out these seminal papers uh so i mean i'm not going to say most of them were uh you know irrelevant but uh as you i mean you can see that yeah they they're they're academic perks but they will not be that impactful or they will not be that's uh remembered in the future so we it's more i mean i mean it's a choice it's a choice you can include them as well it's a choice it's not i guess it's not like is this still a quantitative choice because there are zero citations but it's also quality qualitative i guess yeah yeah i mean probably if we include those the data will be secured in certain senses because uh one of the things that those zero citation papers are not you know equally distributed for institutions uh so there are these certain institutions that have uh more let's say uh uncited uh articles and it is also one of the problems for you know such things like patent analysis as well there are there's this huge number of patterns that have been cited zero times and they have never been used uh for articles as well some people are just writing articles to increase their number of publications and if it is not uh affecting the literature in an impactful manner uh we thought that you know we can just leave that much hey matt that's an interesting question i didn't realize that i mean grover's algorithm was like ignored and then what happened like five ten years later it was rediscovered no this was saying from the start no i i think this was his phd thesis so actually really irrelevant with the algorithm i'm not sure but it's different than the algorithm okay okay thanks i think so i'm not i'm not i'm not very familiar with the history but i hear one of my professors mentioning this in class yeah so and also continue with this like trend of leaving some papers out are these archive papers or like pre-prints or are you just including published papers we only include published papers that have been indexed by web of science so this that they're they're actually so these papers actually passed like a kind of a test right they've been peer reviewed yeah yeah of course of course so let's say you use scopus scope so there are these uh indexes that certain journals can have so you can be indexed you know science citation index is something that never science uses so if you are indexed in science citation index or the social science citation index or the science citations index extended you know things like that you are included in the bible science database so sukupu is actually mrx standard database and google scholar basically includes everything even certain presentations can be included in google scholar so what we did was to take the smallest uh let's say group of papers that we can find that still is within the field so we write in the article that so this is not an exhaustive analysis but we try to kind of do our best to keep irrelevant or non-impactful papers out while keeping most of the seminals works in because it is really hard if you are not going through each paper one by one it is really hard to identify whether an article is really uh should be there so for example one of the reviewers wanted to know you know are you sure that all of these papers are in quantum technologies you know maybe there are some other papers that are related so we went over uh almost if i remembered like two thousands of them we randomly selected two thousand of them and it turned out that there is this thing called qubit fluorometers uh being used in medicine uh fields and it has nothing to do with arguments and there were eight papers on that in the data set even if we tried to limit the uh with certain very specific keywords still there were some uh papers that are uh not very related so it is hard to actually get 50 000 papers that have 100 percent relevancy and all of them are definitely related to quantum technologies but our rate was really high so out of 2000 we only had time on you know irrelevant papers so the we did the ratio analysis in the paper as well okay any other questions about the academic part of this study well so i want to talk about industry a bit um i did say hello uh alexis uh sorry emote either sorry go ahead sorry okay thank you yeah i just wanted to ask that like uh i mean i think zaki showed the trends right like where he like tried to cluster all the countries or institution wise you know and all the domains you know uh the recent trends you know all of that so uh does that like the work published in that right and uh the trends do they match accurately like you know is that can we follow that in that way or like is it like the trends that are there and the work that is being published like you know like you mentioned about quantum teleportation right uh going very high you know at one time you know uh there was a lot of research there and after that there was not much so in that way like do the trends and the work that is being published you know uh can we say that both are in tandem actually uh so if i understood your question correctly you're asking whether these trends show that the research is being conducted right now or it has been conducted and it stopped at some point emma did i understand that correctly because there was this very brief flag and yeah yeah yeah is it fine yeah hello so now i now i can leave yeah yeah so yeah almost that way actually like uh i just wanted to know about the trends you know so now for example if i want to know the current trend like the current a sub domain of quantum computing which is in uh demand do i go towards the papers published or do i look at the work that is being really done on the ground actually that way okay so you can see the list here right okay yeah so and you can so for example here that mean this means that there has been a burst in research in quantum computing like 1998 and it it has faded away in 2009 that means that either the field has become mainstream so it is not bursting anymore and bursting means that the frequency of that keyboard is increasing with respect to increasing other keywords very rapidly so it has been becoming a trend it has been becoming very uh impactful so here and you you can uh see the rest of this list in the paper uh so the end is not uh so these haven't ended or maybe let's say it ended in 2019 uh because that's when our data set stopped so these are the current trends but as you can see the newest one is starting from 2016 which means that there are probably other trends that that have not bursted enough in the literature uh that we can identify them so you are right on both icons i mean if you want to uh let's say catch a uh let's say a tidal wave in the field uh you should definitely look at these trends because they represent a fields let's say uh increasing importance but also you should look at the work that is being done underground because if only let's say a couple of papers are published in a very important topic then it will you will not be able to see them here because this is a very quantitative analysis it looks as at the huge number of papers and it compares them with respect to each other and even if you know we can find a single paper that is very uh you know revolutionary right now you will not be able to see it by this kind of an analysis uh did that did that answer your question yeah definitely i just have a brief a question actually another follow-up that is the whole study in world quantitative modeling right like almost quantitative analysis and you did uh mention about not having like not taking into account the qualitative so in case let's say somebody wants to further take a qualitative analysis of all the of all the papers involved in uh like quantum computing or its sub domains or other this thing so is there any procedure or process that you think of uh that may be taken because that would be too exhaustive to go about for qualitative right yeah i think that's a very good question but uh the answer actually depends on what you are looking for so if you are looking for a way to identify new trends uh there are some methods that use you know the open nlp framework or you can use certain nature language processing models to identify you know what is coming next but as far as i know there are no well-tested methods that are working right now so i think the best way uh to do such an analysis would be to first quantitatively identify the top players in that field and then go have interviews with them or same structured interviews or you can use some questionnaire methods you so there is this method called alpha studies so what do you actually do is you send the questionnaire you get the answers back and then you limit the number of answers that people can give you give them the study you give them the questionnaire back you do you do these like three rounds and in the end you have this very narrow and crystallized answers so that kind of a qualitative analysis would be much uh let's say better and it is much well tested uh compared to these natural language processing methods because i mean you can you know that some of those models produce not so very clear results yeah thank you thanks yeah it might be also like interesting interesting to look at um um preprints right like following up from crown's uh question about identifying current trends you would you would be able to like extract some information from there because there's always a delay otherwise there's always a delay in like getting published and then getting cited so you are actually looking like a a photo from the past when you're yeah yeah exactly uh and i think archive recently opened their data uh base yeah uh but the the problem is this so in archive uh you can let's say you can look at the authors you can look at the title the keywords but uh you cannot check the citation network and citation network actually gives a lot of information um so that needs to be more let's say stream or mainstream or streamline the data set that archive provides because the the only thing that web of science is really good at is organizing the data set so when you download the data set from web of science it is it has very clear distinctions like these are daughters these are deflations these are the references these are the citations so this is the date this is the journal and you know you get this huge uh chunk of data uh so that is the reason why people in bibliometrics use uh that kind of very well structured data sets so as i said you know you can use google scholar as well but google scholar is not that well structured so in the end you will have to clear your data set for quite some time because uh there are many uh problems with certain types of data sets and using preprints is a really good idea and people are trying that for quite some time but the main problem there is structuring your data set properly which takes person hours so you need some dedicated people that probably you need a project well funded for that so that some person can spend you know hours and hours of actually making that data available for the softwares yeah i mean it's a data science problem you just have to read in the papers and connect it's it's a networking problem it's actually a very cool project so if someone's watching this right now and thinking about about work to do in like data science or network graphs go to archive papers and read in the citations and try to connect them together so okay i want to briefly talk about how this can be useful for industry so i did some work for qedc uh the uh quantum economic development consortium uh about uh workforce trends and it was very high level nothing that like this at all but it was just looking at keywords and uh trying to see what what what people are looking for when they're hiring so it might be like these do you think these tools the the web of science tools are are they specific to academic articles or do you think these kinds of data science tools could be used to identify workforce trends in any industry not just quantum computing because it would be very quantum computing is very interesting because it's a very uh dynamic industry it's changing a lot it's explained the hiring is exponentially increasing so do you think this work can be applied to that um well the web of science part cannot be because it is mainly focused on academic publication field but you can use these kind of network analysis or you can use the let's say burst analysis to uh algorithm of claimer uh for such analysis i think people are using those but they are not publishing they are using those internally uh or they are using such methods for you know patent analysis or uh maybe i don't know uh to for capital investment i'm not exactly sure i i think you're right i think uh such methods can be used for you know hiring practices or capital investments uh but i suspect people are already either using those or there is not enough data aggregated to be used because one of the things you need for these kind of network analysis is that you need a lot of paper so you you cannot do such an analysis with let's say 200 papers i mean you can but the clustering would be really terrible i mean you you you will probably get a couple of keywords that are just standing by themselves so if you don't have like one thousand two thousand uh properly uh reach uh let's say data points uh probably uh it will not be rather useful to to do such an analysis yeah i mean again that's a data science problem just to aggregate yeah exactly exactly aggregation part comes first i think uh one one quantum has uh has been working on that or andre has been working on collecting lots of data in the in the quantum industry um okay so if we are out of questions we can uh move on um is there anything else for the key okay great thank you so much zeki this was great very interesting uh we saw that it can be applied to many other uh uh fields uh thanks thanks for the invitation again yeah so nate and i think you want to present something right okay hi nathan are you still with us sorry i was talking to myself i forgot that i'm you okay go ahead i'm snapping to my screen in my cat okay so uh i'll share my screen i think you might find this interesting um hang on uh yes there it is uh okay hopefully you guys see a youtube channel learning quantum computing using misty states yes we see it okay so uh first of all let me back up one of the nice things about this uh group is like you know in the valley we sarcastically call them the three f's family friends and pools when you first start something you know before you show it to somebody you want to show it to the 3s make sure you're not making a fool of yourself because you know you think your work is okay but it's never like that so that's what's nice about a small group like this so i just finished doing this and you might find this interesting it's a diagram a diagrammatic pictorial approach to teaching quantum computing and uh there's a something people might know about uh kirkus book which is a big textbook pictorial programming quantum computing uh but it's a serious you know great uh upper undergraduate and graduate level book so there's another book out that you're looking at here on the screen and the lower left called q is for quantum i'll show you a bigger see if i can make that bigger see it yes okay any guys uh familiar with this book or terry rudolph the name rudolph should ring a bell let me just do a quick test here anybody familiar with uh yes we're familiar with terry yeah okay so he wrote this book as part of uh uh you know he's not just a good scientist he's an entrepreneur uh but he's also interested in educating the public in particular high school so this book has a diagrammatic way he uses let me bring this back hang on here we go okay so it uses this idea of uh these clouds and um to explain quantum computing at least at the fundamental level right so the bottom line is i i fascinated with the book and uh so i put together these youtube videos which you can look at later uh but i just want to show you like perhaps one or a few minutes of one just to get your feedback uh both on his technique and also uh how this is coming across would that work for a minute if you like listening for two minutes sure yeah uh all right so let me uh here um here let's just try this one this is a youtube video inside the youtube video right now foreign it's still spinning you guys see it spinning you see the screen anyway you see a canvas called learning quantum computing data yeah i don't know if it's gonna work but let's see let's see the limits of zoom you can also like put a link if it's not working anything yeah martin i can't hear you i'm saying we can put a link under the video if it doesn't open up okay what are you seeing right now that's what i'm asking it's just loading it's not working right but you're seeing the screen behind it right you're seeing the screen the canvas maybe maybe if you press play it might work i start i tried that i'll do it again hang on there it is next up are classical games you're hearing it now right yeah the great aspect of the campus the rather unexciting logistics of becoming future with its operations but now things start to become a bit more doctorable for example before getting into how each of the five gates works is worth first taking a global view since that broader perspective provides useful insights into both the gates actions and also into how to apply those actions to address a wide variety of real-world conditions observing the five gates at a high level we noticed something interesting they categorized in two ways first according to the action of the gate what it does here we have a knot and swap the details which we'll get to in a moment and secondly second the five gates also segment according to a capacity all gates possess keep in mind this is for high school students beginners action with a control option which determines whether the gate activates or not the result is a more intelligent gate since now it is not applied every time it appears but only if certain other conditions are met you can easily imagine options where you want some flexibility in deciding whether to apply a given gate and that's what the addition of the extra seat hole does in the above five gates obviously the more holds the more conditions a gate has to meet in order to activate making it adaptable to mildly more problems you're likely wondering what's up with the multiple lobes why one or two how about three or even more as we'll be clear in a moment when we divide into the action of the various gates the number of holes is a measure of the gate's adaptability to conform to the problem at hand clearly the more holes available the easier it is to sculpture a specific gate to fit the shape of a problem prior to descending from our high-level view and getting into the mechanics of the gates let us summarize why the division we have highlighted between the action of the gate and its adaptability is beneficial permitting controls to accompany gates as they do in the cnn sea swap and see sea swap cases empowers the originating knot and swap gates in fact in valuing any arbitrary gate with significantly more capacity to model real world situations in summary within these straightforward divisions reside powerful processing capability as you are about to see as we link the high level view and return the gate details all right yeah we're very cool nathan oh seriously it's the first hey you guys who don't know me i'm 73 i'm like retired i'm learning like quantum computing the last three years mark knows that so it's the first time i've done a lot of things in my life but never did something like this i had to learn you i had to learn uh you know how to put up a channel on youtube sites for the video software but anyway overall for a high school person is it interesting is it informative honest opinion you remember you're the three f's uh so radical candor appreciation i think it's super helpful i mean i wish i had we had stuff like this when i was in high school um well no high school kid has seen it yet but uh okay continue um what kind of software are you using is that that's an interesting question camtasia it's because it's like the ms word of uh uh video editing software but it's a legacy package goes back to you know early window days and it's a steep learning curve if you love a better like my intuition the first thing i thought about when i saw the uh diagram was that i want to play with it like i want to drag gates around oh yeah yeah oh for that's different uh that's a jscript it's a it's a open source canvas right there's other videos that let you do that and it's got it's going to be online so i'll send something out you guys are uh mostly guys around my list that i'll send out uh yeah like if these students not but like uh in addition to these like kind of mini lectures they could like build their own here let me show you hang on that's a good question sorry uh here you see it now yes so i can take a gate this is i had somebody help me watch i can drag a gate down put a ball on top of it right it doesn't quite land right but you know it's a work in progress and then it's a nut gate so what do you expect to happen it's going to embarrass me to come out black but let's take a look right maybe i get lucky and you go to run and lo and behold it came out white nice all right so anyway we build up you know uh here's a swap look lo and behold it's swapped hey cool i mean you gotta remember it for uh i'm not gonna get people fascinated and interested you know and then you can clear it you know and do other things we're working on it but i'm trying to terry's hopefully going to help us with this right so uh he hasn't yet seen this so this is behind the scenes those videos which are are intended to like give you an overview it's an intro to the uh to that canvas and some background you know uh and so what it is those are powerpoint you know it all those animations those uh items that come up those are done in powerpoint you make a png out of them and the video software imports them then you turn on its virtual screen and record it but you know figure out all the logistics of that is ordeal so now that i'm i'm doing it it's going to hopefully be easier anyway you think it's useful right yes that's all right let me let me uh just so you can keep up and give me more feedback let me figure out why uh where i can put a chat in here on i've got the wrong uh as usual with uh zoom um all i see is you guys faces how do i get back to uh where i can put in chats uh maybe if you stop sharing that might help oh yeah that might be it okay all right yeah uh there's more but let me give you the uh the address where you can get to it all right and then look at it and then get back to me by email but yeah glad to show it to you hang on let me stop sharing that's a good point yeah that worked okay i'll put it in the um in the chat great okay uh thanks for sharing nathan yeah you need to type it in before you hang up okay yeah yeah yeah i won't i'm waiting um okay so um i guess we can wrap up um thank you so thank you is a key for uh presenting today also nathan thank you for your like lightning talk um it's great again like right sorry we are ranging from like high school education to phd level research to in industry applications to hot uh workforce trends so our our talks include many many different uh many different fields and we will we usually do like them at the end of every month on wednesday so we will we will have a few talks coming up um we'll keep you posted our website is quantum dot sv and yeah nathan just shared uh the players that's on the chat um thank you all for coming all right everyone see you next time recording zeki thanks for the talk thank you