Adding a Federated Search box to the Library’s homepage

We would like to suggest adding a search function for 360 Search (federated searching) to the Library’s home page. We feel this would be valuable, both because we are receiving positive feedback from people who have used it, and to make it more obvious so that it gets more usage and we get sufficient feedback to evaluate its usefulness.

We feel that this would not be making a major functional change to the home page – we are adding functionality and not changing or deleting existing material. A search like this was discussed at the time we were considering the new home page and it was only left off at the beginning of the year because at that stage it was felt that 360 search was not working well enough- since then the major issue of date sorting has been resolved. The best time make a change to the Library’s home page would be over the mid-semester break.

To see how such a search function would look on the home page, have a look at the draft page .

Please give us your opinions on this suggestion, added as comments added to this posting. We will review them and make any changes needed as discussion continues. A final decision needs to be made by the end of June so that we can put up the new page at least a week ahead of the start of term.

Catherine Jane and Caroline Syddall

19 thoughts on “Adding a Federated Search box to the Library’s homepage”

  1. The Quick Search on one box and Easy Search on the other may give an odd impression. I would like to suggest Database Cross Search or Database MultiSearch or Search Multiple Databases or something along those lines that better states the purpose of the search box

  2. In principle it is a good idea. In practice it is a really bad idea for one reason only – speed. I just did a few searches and the AVERAGE response time was 19 seconds before anything was returned. This was way way way too slow! Even 10 seconds is too slow. If something can not be done about the response time I do not think we should be using federated searching particularly when there may be other options that do a good enough job – http://www.scirus.com, scholar.google.com etc

  3. and (did i mention) if something doesn’t work or goes way too slow it is a really bad advert for the library – It makes us look like old fuddy duddies who tolerate really slow stuff and hang out in all the wrong places and never get invited to the right parties etc etc. BTW – it doesn’t seem to work on my mobile phone either….

  4. Can I ask Adam what "it doesn’t work on my mobile phone either" means? … you can’t load the homepage, … can load the homepage but the 360 search doesn’t work, … something else altogether

  5. The speed issue still bothers me quite a bit I’m afraid – although I thought that the 19 sec average for Adam was fairly sharp for a Fed search product! Web of Science just took 16 sec to load an initial search page (longer than for the 360 search screen), which also seems way too slow. Google Scholar’s search page loads in around a second and search results returned in 0.29 second, which I guess is the new benchmark
    Which all makes me think that putting 360 ‘out front’ for a while might be no worse than some of the other comparatively sluggish performers we have on our databases page. Maybe that would be a compromise – highlighting 360 via the http://library.canterbury.a
    page?

  6. The nice but slow library Easy Search.

    MMM What can I say. Nice Basic search of "Dubai Towers"
    All Databases (default choice) = 62 seconds.
    Quick Search (*cough*) = 47 seconds.
    Engineering = 53 seconds.

    At the least, the dropdown menu should default to "Choose a subject", and it might be better not to have "All databases" as an option.

    It might be worth putting a hyperlinked "?" by each of the catalogue and database search titles, linking to pop-ups which explain what each of these are for — a lot of students are not clear on the distinction between the two things.

    A warning that the search won’t be instant would be useful too eg "(will take up to 60 seconds)" after the title.

  7. Another thought:
    On both the catalogue and databases search boxes there should be a "(More [catalogue|databases] options)" link going respectively to the main catalogue and databases pages – rather than giving the impression that this is all there is.

    I do think it’s a very useful thing to highlight on the front page, as long as there’s sufficient information available for students to understand what to expect from it.

    If for whatever reason it doesn’t go on the front page, it should at least go on the databases page.

  8. I think we should just go for it on the homepage, and then survey users.

    I talked to user just after reading the comments on speed this morning. He was said that as a user he does not know the appropriate databases, and 40-50 seconds is a lot faster than trying to find the right databases and then using each one. He has managed to find 360 search and likes it.

    We have 9 databases in the quick search. Each should be reviewed for importance and speed. It has now been moved to the top.

    We should not promise to improve the speed as it is unlikely to happen, there are too many variables in this type of searching.

    This is a relatively expensive product. It needs an opportunity to stand on its own merits, otherwise it needs to be cancelled. I think that would be a pity given some of the comments from the currently small number of users.

  9. Thanks Catherine and Caroline for drafting this. I think the fact you both feel "happy" to suggest this confirms my feeling we should do it. We could get very bogged down by the all the detrimental consequences including speed. But we need to explore the possibilities and the benefits and if we don’t promote it to users in an up front manner we’ll not be able to do this. We can always discuss the decision again if it turns out peoples lives (or the library’s) are (is) in grave danger through the introduction of such a terrible concept. I think Deirdre has a valid point "multiple" or "multisearch" is possibly a little less misleading than "easy" (although we could spend forever angsting over the perfect terminology) and have just one other suggestion. Maybe in small font there could be a little link "What is this" next to the text "Databases Easy (or other wording) Search" so people can click it to get an explanation, or link to http://library.canterbury.a… I’ve just logged a fresh report of the slow searching/major irritating blank white screen and also asked Serials Solutions if there are ways of making it more obvious to people the search is processing (something is at least happening).

  10. Whilst I acknowledge that the search is slow, I don’t think this means we should give up on it.
    I’ve had a number of users who, as well as being very pleased with the results they got, were quite excited to be able to identify useful databases through this search.
    I’ve also had some students who would not have the patience or confidence to search several databases separately who have found useful resources- resources which they would not have found through Google Scholar, or by searching a single database.
    360 won’t suit all (and we were clear about that from the start) but it does suti some. I think we need to give it a go- and make it visible enough to hear what our users think of it.

  11. Response to Catherine: only the 360 bit does not work

    Response to CarolineS: Can you give us some examples? This may be a science versus non-science thing.

  12. I’ve made some changes to the proposed look of the page, in response to the above comments. I’ve changed the name of the 360 search – thanks Deirdre – and I’ve expanded the titles so that if you leave your mouse over the headings you get some explanatory text coming up. I’ve also added a small question mark in each box. This has the same mouseover text, but also links to a page with more explanation. At this stage the explanations are very limited – I need some more content for these.

  13. I support having the 360 search on homepage. It was way too buried before to give it a decent chance. Having it visible will provide more feedback.

  14. I too support the 360 search on the homepage (but perhapes called something less dizzying). I think it is exactly the kind of access point that users are looking for in the google age.

  15. I don’t think we should be treating speed as an optional extra. Speed and ease of use are the minimum expectation of our users. If it is not fast enough we should not be paying for it as they won’t use it. This is particularly the case when there are other things out there that are free and that may in fact do a better job. Scirus & Google Scholar equally deserve being promoted to the Library Home page. Do we have any evidence that they do not already search much of what we are offering through 360 and more (only a lot faster)? Perhaps we could use 360 to focus more on materials that are not indexed elsewhere – e.g. New Zealand materials or UC materials?

    “Truth always rests with the minority” ~ Kierkegaard

    😉

  16. Even though the search isn’t speedy, it’s a lot quicker than running through multiple searches.

    Scirus is really just science-only. http://scirus.com/srsapp/ab

    Google Scholar is much improved since they’ve added Elsevier content, but they’ve never been open about what they index so it’s hard to know without running brute force comparisons. I think they’re pretty good in sci/tech but I don’t know about other fields. Its relevance ranking is bizarre too. I do think it’s worth promoting it more on the databases page but I’m not sure whether or not it’s an overall improvement over 360, even taking speed into account.

    Catherine: The changes look good, thanks! Giving some thought to wording/explanation…

  17. Scirus is definitely dominated by STM materials but an eye needs to be kept on it given that it also includes growing amounts of open access materials.

  18. There are definitely some problems with Google Scholar — try searching Nanotechnology articles published between 1910 and 1930!

    However that being said there are increasingly articles around like this one showing GS is VERY competitive in non-STM subject areas also – http://dx.doi.org/10.1016/j

    If users are not looking for a specific article do they care whether or not a particular publisher is covered? If they have easy links to full text do they care whether or not GS gives them an abstract?

Leave a Reply

Your email address will not be published. Required fields are marked *