Ugh! What a horrible UI revision. Mobile has completely destroyed web design for anything larger than an iPhone. Stupid round pictures? Really? (And thousands of clean square ones now vignetted and made to look stupid against a round background?) Items in the Areas of Interest list chopped off with a ... and so made illegible? More empty white space pushing actual data off the bottom of the screen? The Metrics section used to be extremely interesting, with comparative data from many fields; that looks to be completely gone -- how could something that useful be eliminated? [Wait, the "learn more" page says the subcategories are still there in the left column, but I don't see any left column.] Come on, Google. Don't any of your people use actual computers any more?
Agreed completely. More insulting is the fact that this is dumbing-down of Google Scholar, and so presumably aimed at a demographic of above-average intelligence and emphasis on content compared to the typical computer user.
"Cite" and "Save" are self-explanatory, icons are not. They're also unusual on a line with plenty of other text (which no doubt they would now decide to "iconify" in the future...)
I've seen enough of this trend that phrases like "cleaner look" make me think of an empty room left behind by someone who decided to throw away everything you need --- bland and useless. It's beyond stupid.
I suppose we should be overjoyed anyway. This update means that Google is not shutting Scholar down - at the very least, not today. Knowing their habit of axing products that do not quite attain civilization-level scale, it's always a grim possibility I keep in mind.
In Google Scholar - just a few days ago - it used to highlight whatever search term you used on the main page. Did this change with the new version? Because when I search "case law" now, when I go into a particular case, the search term is no longer highlighted.
Meh, I just looked at Scholar for the first time in ages (since I am no longer a scholar). The information density looks as high as it always has been -- even if they are claiming to have the dreaded "cleaner design".
I agree that text is better than an icons, especlally the misleading "quote" icon that actually means "cite" -- so thank heavens for tooltips (which presumably work only on desktop).
What is awesome is that there are these complaints to begin with. While I was still in academia, I'd have loved to have access to this sort of thing.
We could search back then. It was a paid service, involved a human who actually did the searches as a service, and took several days to get you your results. You did contact them online, so there's that!
But, today, a gripe is needing to hover over an icon to read the tooltip. That is truly awesome, in every sense of the word.
As an aside, when people say that the Internet was better back before commercialization, they are telling horrible lies. No, no it was not. It was pretty horrible and computers pretty much sucked. The hardware, network, and content are much better today.
In 2017, I don't understand why UIs are not totally configurable by now. It should be clear that people have different tastes. I am a fan of full-text dense interfaces a la Reddit but I know I am in a minority. I wish I could activate a content-only CSS.
Well, I do understand actually. It is about ads revenue. Fuck that. I wish this ecosystem finally dies.
It's not about ad revenue, it's about people almost never taking the time to configure things they can configure, and that if you don't use configuration options, they are distracting/can cause bugs.
It should be configured in the browser: I want to display news items this way, I want my image galleries that way, I want my video player with that llama interface...
Start paying for software then. (Not just directed towards you, but also towards anyone who has these thoughts. That's the only way we get off this model, by showing other models are profitable as well)
You can use userscripts and CSS in the browser, and/or rewriting MITM proxies (when the browser is not as featureful or easy to control), although personally the combination of usefulness and annoyance has to be quite high before I start trying to restyle a site --- not very many qualify.
CSS completely off also tends to produce good results, but usually not for "app-ish" pages.
Google Scholar isn't the worst case of it, but Material Design is making a lot of the Web very unpleasant. Over-saturated, garish colors, low-contrast text, and too much animation (not accessible to people who have trouble with visual motion). Round profile pictures are also not a good trend -- users' images should never be edited without their permission, because some of them depend on being square. MD could be fixed by toning it down a little. (Ask "how can I minimize the animation and colors as much as possible?" No delays, sudden fast movement, or movement from areas of the page that don't need immediate attention.)
Scholar is by far the most powerful search site for academic literature. They not only find new publications weeks before other platforms do, but also find far more literature from all sources.
Nevertheless, I feel like this is one of these half-hearted Google projects and doesn't fully live up to its potential. This relatively subtle UI change is a good example that they don't seem to care too much about taking it further. I just pray Google doesn't shut it down one day.
Its interesting how the big guys never truly paid attention to their academic search portals. I remember MS Academic Search to be very useful, which thereafter remained as is to the point of being clearly outdated. And then, on a bright sunny day, discontinued. Only to emerge years later, imbued with a fresh interface, horrible search and confusing UI.
Scholar too felt ignored like its a bastard child, but this new update looks promising, not in term of their changes, but due to the fact that this implies they haven't boxed it yet. Also, they haven't "revamped" the search, so I'm not worried about the core functionality.
Most likely this behavior stems from the fact that this service doesn't generate revenues for them (I'm assuming so, please correct me if I'm wrong). If so, then how does a small company keep up its own academic search portal?: Semantic Scholar by Allen.
> this service doesn't generate revenues for them (I'm assuming so, please correct me if I'm wrong).
Considering that there are quite a few researchers working at Google, I can imagine that the fact that they have this tool available to them indirectly causes some revenue. (And also, goodwill in the academic community, i.e. a community of potential future employees.)
Semantic Scholar is completely non-monetized (and will stay that way as far as I can tell) because Paul Allen pays the bills and has a public interest goal.
If anyone from Google Scholar is reading this, please, please, add a DOI field to the BiBTex citation. It is required in every serious journal and it's tiresome to copy the citation, then open the article page and look for the DOI to add manually.
This small change would be more welcome that any other UI refreshment, I guarantee it.
Wow, both Alex and Anurag are still working on Scholar. They started the project almost fifteen years ago. Google infrastructure all around it changed entirely over that time; I'm wondering if this is another major rewrite or "just" a substantial UI update to meet Material Design guidelines.
It seems just like a UI tweak. Which is fine...but anyone who is doing serious research or is just an enamored amateur will quickly get over the quirks of their tools.
Most of the rewrites are behind the scenes. You'd only notice if you looked e.g. at latency, reliability and trigger rates. I think Scholar is even older than Borg.
Please please please fix "sort by date" filter. Searching for "natural language processing" returns 3,330,000 results with the default "sort by relevance" and 4,730 with "sort by date."
I've noticed this completely nonsensical behaviour with regular Google too. "Sort by" shouldn't ever be a "filter" --- but an ordering.
The result counts are usually highly inflated and approximate anyway (i.e. can you eventually actually get all 3M+ results?), so perhaps they don't mean all that much.
The best explanation for that anomaly, which is more of an excuse, is that they don't want to spend computing power sorting all the results.
Nevermind the fact that sorting by date is probably much easier than computing "relevance" and sorting by that...
Only guessing, but maybe they don't have a date for the rest. In that case one of the sensible things to do is what they seem to be doing - only showing those for which they do have a date.
In my experience the quality of the BibTeX offered by Google Scholar is really low, too much missing or incorrect information. Crossref is much better: https://www.crossref.org/
The change I really wish Scholar would announce is adding an official API. I'd like to be able to link papers in my library back to some kind of Scholar ID and walk out the citation and 'related' articles from there. I'd feel much better if the only option wasn't a screen scraping library.
It's really a shame that no university library consortium has tried to build this themselves.
I wish there were an easy way to browse the most recent papers by category. Or get an email of the daily digest of what's added today for my favorite categories.
I know arXiv does this, but it's in such an unreadable format that it makes me just want to read it on the site.
1. It's not a simply algorithm, but yes, page rank plays a role (using citations rather than, or in addition to, hyperlinks).
2. crossref.org has an API? Or do you want the actual papers included - because if so, open access needs to see significantly more uptake first.
"Author profile pages got a cleaner look, especially on mobile devices. Rest assured, we did not change your citation counts - at least not intentionally."
That's funny, because I was just thinking how Academia.edu has become a joke site, while everyone in my field uses Google Scholar daily. It's almost impossible to find references, put together a bib list, or navigate through related work without Google Scholar.
Instead of becoming a more or less of a clone of Facebook news feed for researchers, RG has become infuriatingly full of nag boxes, and in its efforts to get you use the website more prevents you from actually using it at all.
Some are useful. But it has begun to cross the line. For example, the nags for their new feature for organizing research into projects are annoying, and then I started getting notices that a deadline was approaching for an update to my project. Eff that. I have plenty of real deadlines already.
I never use academia.edu unless it shows up in the search results. I see no advantage to posting my work on their web site when I could post it on my own site just as easily.
