Peter Ingwersen na KISKu: velký rozhovor

Peter Ingwersen je dánský vědec, který je světovou veličinou v problematice Information Retrieval - vyhledávání informací. Je reprezentantem nového proudu v tomto oboru, který v sobě spojuje poznatky z výzkumů v IT a ve společenských vědách. Jak se dívá na budoucnost knihoven a informačních specialistů? O tom a mnoho dalším se dočtete v následujícím exkluzivním rozhovoru:

Mr. Ingwersen, how do you like Brno? How do you like Czech Republic?

This is the first time I'm in Brno and the Czech Republic. I'm happy for this invitation and I appreciate very much to be here. So far as I can see from the two visits in restaurants I've been to, people in Brno are very friendly and they like the same things as I do - beer and sausages and things like that...

Have you ever been to Czech Republic before?

No, never before. I've never had the opportunity.

What do you like about Prague?

The architecture is fantastic, and the way the city is situated reminds me of Budapest, but Prague has much more beautiful old part of the city. Pest was also very much destroyed during the war, whereas here are older buildings much better maintained. I appreciate it very much, especially these narrow streets in the old city of Prague.

Let's be honest, you're a big star in a small world of information science and among information specialists. How does a man become who you are?

I think it has to do with combination of timing, will - in sense of you want to do something - and also opportunity. I was very happy to have very good teachers back in seventies in Royal School of Library and Information Science. They pushed me into initial research projects, and then I was lucky again that I had the chance to go to European Space Agency in Frascati for three years. There I readied on how to make research because it was full of physicists, mathematicians and biologists. My first articles, which came out while I was there. I think that fundamental basics of my career as an information scientist started there. Another important thing is that I come from small country and if you come from a small country, you don't have to compete with so many other people in your own cultural scientific sector. The real competition starts when you try to become international. This means you are better off in smaller countries like Denmark, Czech Republic or Sweden. If you want to expand outside and be international, you have a better starting chance, because people don't see so many Danes or Czechs or Swedes in the international scene. That is full of UK and US people.

So it's definitely an advantage for a scientist to come from a small country?

Yes, if you have the right skills and knowledge and willingness to participate, of course.

You cannot criticize computer scientists if you don't know anything about algorithms. So you have to learn that. That was what I've done since the eighties. 

How many universities do you teach at?

Right now, I'm Professor Emeritus at Royal School of Library and Information Science in Copenhagen. Then I'm part-time professor in Oslo University College, where I teach scientometrics and research evaluation. Then I'm docent at Åbo University in Finland, where I teach twice every year.

How would you compare the universities?

The Royal School has always been very big, covering cultural aspects of library, information science, library role in society, plus Information Retrieval, information seeking,  research evaluation and citation analysis. Whereas Åbo is very much about information management and not very much about Information Retrieval. So, when I'm there, I'm talking to the PhD students and Master students about Information Retrieval or scientometrics. They are also starting now in webometrics. Oslo is much more about seeking and retrieval but they didn't have anything about scientometrics so that's why I teach scientometrics there. I've also collaborated very much with Tampere University, where I've written several articles and a monograph about Information Retrieval. I would say Tampere is the best place in Scandinavia for Information Retrieval and daily life information seeking. They are very strong in these aspects.

Computer scientists have no real idea about social science methodologies. On the other hand, they have lots of good ideas and are much better mathematically equipped. 

You represented a user-oriented approach to Information Retrieval. How would you describe it?

Now I'm representing an integrated approach - integrating user oriented cognitive interpretation stuff with hardcore Information Retrieval.

But you started in the user-oriented wing, is that true? Did you have any misunderstandings with people from classic IT approach to Information Retrieval?

Yes, I started in the user-oriented wing. Well, if you're criticizing the other guys - the mainstream Information Retrieval approach - then you have to know about it. You cannot criticize if you don't know about algorithms and so on, so you have to learn that. That was what I've done since the eighties - I've been trying to reinforce my knowledge in the laboratory environment.

What were the relationships between people of the user-oriented and the classic approach?

I think it's very natural that they are different. They come from two different research schools. The first - library and information scientists and so on - they are fundamentally of social science and humanities, so their work is basically about qualitative research methods. And the others, computer scientists, they have no real idea about social science methodologies. On the other hand, they have lots of good ideas and knowledge about how to make experiments with IT and they're much better mathematically equipped from the start, because they come from computer science. Social scientists don't know anything about mathematics, algorithms and so on. That's two different schools, that's why we now try to put them together. We have a conference which we started in 2006, it is called Information interaction in context. That's really to make a bridge between younger researchers from computer science who'd like to do information interaction and then softer information scientists who'd like to do it from more human perspective. Now we had third conference last year and we'll have fourth one in Holland next year.

Is it a challenge when you stand between those two sides?

No, I don't think so. It's nice, if you admit that you'll learn something you don't know (laugh). In order to be respected by those sides, you have to respect both sides. You have to know their limitations, what are the challenges and advantages on both sides. That's why I like to teach people who are good at stuff on both sides. It seems to sbe successful to certain extent, so I'm happy for that.

There have been several investigations looking if PageRank is a good search engine from topical point of view. The results are not really positive, in essence.

You invented the word "webometrics". What's it all about?

Webometrics is when you take methods of bibliometrics and do in on the web. You do same things that you do in scientometrics: you take a space on the web, for example an academic space, where you try to find out which networks are built up, how can you measure the impacts of the varies sites in this network and you also would like to know: is it a difference if you search on Google compared to Yahoo? Are they building two different networks on the same space, simply because they cover different things? These coverage aspects of search engines are also part of webometrics, aside from pure link analysis. You can also make a citation analysis on the web: you compare academic sites by inlinks to the real impact by citations of their publications lying on their websites. There are two different measurements, where one is webometrics - the link analysis - and the other one is some overlayer with scientometrics, because you "measure science" in web environment.

What about PageRank? You criticized it for having no real connection to relevance of the pages from search results.

No, PageRank is fine for Google and Yahoo and Bing and so on. It is based on some kind of weighting over inlinks and other sites' pageranks. But as far as I can see, they have to move more and more into Information Retrieval related weighting schemes based on content, because they get too many no-good results out of just using links. I assume that pageranking will more and more use higher weights on content-bearing features and less on the links. Because, what we know from webometrics, it's the commercial sites which have most inlinks, so basically if you put lot of emphasis on them, commercial sites will always win in the ranking, which might not be a good idea all the time. If you're searching for buying some tickets, then it's fine, because it's well known item - PageRank and alike are very good for that. But if you're searching for topics, then it's no advantage. There's no real connection between links and relevance. There's connection between links and recognition, links and navigation but not necessarily a connection between links and topical or situational relevance. It's perhaps the same as with citations. There have been several investigations of links and PageRank looking if PageRank is a good search engine from topical point of view. The results are not really positive, in essence.

Each public library has no real necessity for knowledge about indexing and cataloguing any more, which is not smart. At the end, they might be the losers.

Do you see any trends of Information Retrieval in institutional databases?

I think there is a trend at university libraries. More of them are using commercionalized ranking systems which are not necessarily Google- or search engine-like. This means that they're based on Vector space or probabilistic era engines and therefore much more close to traditional Information Retrieval, which succeeded twenty years ago. We have been part of testing it also. It's quite interesting, because in library environment you have a problem that you have to integrate many different media, many different document types. Because of tradition they've always been indexed very differently, so what you have now in the library catalogue is some kind of fulltext mixed with a very poor catalogue cards like records, table of contents of monographs. I see a trend that bigger libraries are trying to cope with the ranking algorithms in Google. I'm also suspecting that at the end the losers might be the public libraries, especially in the countries like Denmark, where all things in public libraries are extremely centralized. The cataloguing and indexing is done in central organism and then distributed, so each public library has no real necessity for knowledge about indexing and cataloguing any more, which is perhaps not so smart.

More students should concentrate on human-computer interface, instead of staying on the lines of Microsoft, Apple and Google. They're very limited in their way.

Do you also concentrate on Human-computer interface?

Not anymore, but I encourage that people are doing it. I also work together with some known persons in Danish University computer science department, who are doing interface utility and usability studies. I would like if more students would do it because I think it is better that we get some smarter ideas about how interfaces should be done instead of that we are on the lines of Microsoft, Apple, Google modus of interface. And they're very limited in their way... In fact, I like Yahoo's interface more than Google's interface, especially with integrated searching.

How do you see the role of information specialist in the future?

The role of information specialist has changed radically since the mid-nineties because of the personal computer and the internet. As far as I can see, the information specialist is very usable if he or she works for example in industry or organizations and of course libraries. It would be a good idea if information specialists were the supporters of end users in work situations, because it's easier and more effective to learn how to deal with information in specific domain than doing it on a generalist domain, where you can only scratch the surface. I think end users can easier do it themselves. Why should they go to the public library? They will go there if they want to search information in books, but for gathering information? They will never do it in library in the future, why move there when you have the internet just at your fingertips? Some librarians and information specialists have to change their role radically.

If you close yourself into your culture and language, then basically you are about to be put in the second class science, because you don't really know what is going on.

Do you see any trends in monitoring and marketing on social networks?

As far as I can see, Facebook, LinkedIn, blogs and other social media try to make a commercial benefit out of themselves in terms of marketing and information management. For example Facebook can be used as marketing tool for big companies instead of big companies having their own website on the net. The trend is that they should have only their website on Facebook, because the marketing value of Facebook is bigger than on the net itself - this might be true or not true, that depends. Facebook or LinkedIn, they both have kinds of different characteristics of their users. Of course, they might be a platform for marketing, for certain kinds of industry and private enterprise. But in the end I still think it only depends on who's going to pay for what on the net. There's also another trend, that you have to pay for everything which is worth to get at. Many more databases are still in the hidden web than in real web and you have to pay for going into those databases. Then some databases have gone the opposite way becoming free of fees. I think it will be interesting to see one thing: What will happen to real qualitative data? Will they still be fee-based or will they also try to come on an open access model

Thank you for interview. Do you have any more visions to add?

I hope you guys - especially new students and young researchers - will start writing about  information science much more in the English language. This field has a lot of different journals for different purposes with different levels of access and acceptance so I think there're are many possibilities to mix into the international society. Every time this happened for people in small countries like Norway, Finland, Denmark, Sweden, Estonia, they really benefited from it. Whereas if you close yourself into your culture and language, then basically you're about to be put in the second class science, because you don't know really what is going on. And if you don't mix up with people and go to conferences and try to write into journals, then you'll not be known. And it's better when people know you and you are part of a network and therefore you can get something from people. Thanks to knowledge and accessible technology you can now get it much easier what is going on. So that's my wish for you guys, you should try and go for it!

Fotogalerie

Líbil se vám článek?
Stáhnout článek v PDF

0 komentářů

Přidat komentář

(If you're a human, don't change the following field)
Your first name.
(If you're a human, don't change the following field)
Your first name.

Přečtěte si také

Přihlášení Registrace
RSS Facebook Twitter YouTube
Zobrazit standardní verzi webu

Taky děláme

Feedback