3 ms·

Can't websites like Google Scholar help here? There should be some algorithmic way to identify such journals. One way could be to use PageRank for a paper to cr

by sytelus 10y ago

Can't websites like Google Scholar help here? There should be some algorithmic way to identify such journals. One way could be to use PageRank for a paper to create a PargeRank of journals. Then take the bottom of journals and use either wisdom of crowd or human editors to de-list them or flag them. If publications from bad journals stops showing up (or at least get flagged) in search results and/or Google Scholar, they practically don't exist.

jonathanstrange 10y ago
This would to more harm than good. Using automated methods to evaluate science can only make things worse. Hiring committees and funding authorities already rely on way too many automatic indicators such as impact factors and citation indices anyway, despite the fact that nobody has ever been called a great scientist because he or she published a lot, and, even worse than that, in the humanities people often get quoted a lot and make a career because they published a really bad article that somehow made it into a good journal. Everybody tries to correct them, they publish a few even more outraging and stubborn follow-ups and pronto - they've got their tenure. I've seen that more than once. Besides, whether journals show up in search results or not is also not that important. This is important to find readers and followers, but what counts more for many individual researchers nowadays is whether or not publications show up in their CV. To fix the problem, universities and funding authorities must return to putting a higher emphasis on the scientific evaluation of the actual work of scientists, evaluate their results, their approaches and how promising their line of research is, instead of bean counting. There is no other way to ensure quality other than evaluating on the basis of quality rather than quantity.
dragandj 10y ago
That is irrelevant. Anyone who is remotely competent to read any scientific article will recognize predatory journal with a quick glance. I doubt that predatory articles have even a couple of reads. The problem is something else: those articles are used to beef up the CV of the "scientists" who are then promoted on their jobs over honest candidates. That may be rare in USA (but I am not sure) but it is rampant in developing countries (of that I am sure).
ajdlinux 10y ago
The "PageRank of journals" sounds suspiciously close to the Impact Factor that so many academics really really love...
denzil_correa 10y ago
Google Scholar already has something called Metrics [0]. You can basically go through any field and its sub category to get ranked lists based on h5-index and h5-median. For example, this is a list of top ranked conferences for Databases & Information Systems via the h5 index [1]. [0] https://scholar.google.de/intl/en/scholar/metrics.html https://scholar.google.de/intl/en/scholar/metrics.html [1] https://scholar.google.de/citations?view_op=top_venues&hl=en&vq=eng_databasesinformationsystems https://scholar.google.de/citations?view_op=top_venues&hl=en...
Someone 10y ago
PageRank doesn't work. With both scientific publications and web sites, those wanting to game the system will simply create lots of page/publications that link a lot to each other. That's doable because it is very cheap to produce a zillion web sites/journals and populate them with 'content'/'scientific papers'. In the end, somebody has to make a judgment call as to what a link is worth.

by sytelus 10y ago

jonathanstrange 10y ago

This would to more harm than good. Using automated methods to evaluate science can only make things worse. Hiring committees and funding authorities already rely on way too many automatic indicators such as impact factors and citation indices anyway, despite the fact that nobody has ever been called a great scientist because he or she published a lot, and, even worse than that, in the humanities people often get quoted a lot and make a career because they published a really bad article that somehow made it into a good journal. Everybody tries to correct them, they publish a few even more outraging and stubborn follow-ups and pronto - they've got their tenure. I've seen that more than once. Besides, whether journals show up in search results or not is also not that important. This is important to find readers and followers, but what counts more for many individual researchers nowadays is whether or not publications show up in their CV. To fix the problem, universities and funding authorities must return to putting a higher emphasis on the scientific evaluation of the actual work of scientists, evaluate their results, their approaches and how promising their line of research is, instead of bean counting. There is no other way to ensure quality other than evaluating on the basis of quality rather than quantity.

dragandj 10y ago

That is irrelevant. Anyone who is remotely competent to read any scientific article will recognize predatory journal with a quick glance. I doubt that predatory articles have even a couple of reads. The problem is something else: those articles are used to beef up the CV of the "scientists" who are then promoted on their jobs over honest candidates. That may be rare in USA (but I am not sure) but it is rampant in developing countries (of that I am sure).

ajdlinux 10y ago

The "PageRank of journals" sounds suspiciously close to the Impact Factor that so many academics really really love...

denzil_correa 10y ago

Google Scholar already has something called Metrics [0]. You can basically go through any field and its sub category to get ranked lists based on h5-index and h5-median. For example, this is a list of top ranked conferences for Databases & Information Systems via the h5 index [1]. [0] https://scholar.google.de/intl/en/scholar/metrics.html https://scholar.google.de/intl/en/scholar/metrics.html [1] https://scholar.google.de/citations?view_op=top_venues&hl=en&vq=eng_databasesinformationsystems https://scholar.google.de/citations?view_op=top_venues&hl=en...

Someone 10y ago

PageRank doesn't work. With both scientific publications and web sites, those wanting to game the system will simply create lots of page/publications that link a lot to each other. That's doable because it is very cheap to produce a zillion web sites/journals and populate them with 'content'/'scientific papers'. In the end, somebody has to make a judgment call as to what a link is worth.