Google Scholar Search

I’ve tested a Google Scholar search both using Safari and DEVONagent. My search entry in both is Polyphony AND (Stutter* OR Stammer*). I’m not sure if all of those operators work, but in any case, I used them in both.

Some returns were identical (as expected) and DEVONagent returned less results (also as expected, since I know it is doing some upfront filtering work for me). What’s curious, however, is that there are many relevant hits using the normal Google Scholar search in my Safari browser, that do not show up in the results or (as far as I can tell) the digest of my DEVONagent search. I have the DEVONagent search set to not follow links, filter archived pages only, No Scanner, and to search in Title, URL, Description, Text, and Keywords.

Any idea why this might be happening? I want to use DEVONagent for my research-specific searches, but I don’t want to miss relevant things in the process. Thanks!

EDIT: I’m also running into the issue that when I hover over “data” in an attempt to “add all to archive,” I only have the option to “add to archive” and it adds the most recent hit I’ve clicked on. When I go to the help menu and search for the “add all to archive” option it highlights the exact place where “add to archive” was and still is if I try to access it without first using the help search(???)

EDIT: Apologies for the continued updates to this query, but I’ve looked at the log and noticed some of the hits I want that show up in the “normal” google scholar search. I then noticed that, in at least one example, the google scholar search on Safari must be drawing from the pdf/article text itself to locate my search terms, whereas DEVONagent limited the extent to which it searched the page/document. Is that easily fixed with a setting change?

Another example of this/a similar issue: I created a search set so that I could predetermine the settings rather than editing them after the initial search. I have it set to search Google Scholar, as before. I searched Breath AND Stuttering. Several hits were considered “No Match” and I had to scour the log to find these. For instance, I located the Google Scholar search page in the log and then looked for the first hit on that page to find “A Point of View about Stuttering,” an article which was found but also discarded to the log as a No Match. That article includes in the content of the search both of my keywords: Breath and Stuttering. Why is not being provided as a result? I think it’s highly likely this is user issue, but if it isn’t, and this is working as intended, I find the use and credibility of using DEVONagent as a research tool suspect. Please help!

DEVONagent matches the downloaded pages/documents on its own. You could use the secondary query * if you prefer unfiltered results. But using DEVONagent as a frontend for Google is neither a recommended usage scenario nor does it take advantage of DEVONagent’s features.

To be clear, what I’ve described above is DEVONagent working as intended then? And, am I really using it as a frontend for Google when I’m using the built-in Google Scholar plugin? Is DEVONagent, by design, supposed to be filtering out very relevant resources for my research that a regular Google Scholar search does not?

By default DEVONagent matches pages on its own, it does not only present results of a search engine in a different user interface. This is e.g. explained in the chapters Read Me and Getting Started of the help.

I am trying to “search multiple search engines and websites with just one click.” I would like to “keep-up-to-date on a subject.” How can I carry out these use-cases for which DEVONagent is intended if I am not able to have it return thorough and comprehensive results on a topic or subject? Did you have a chance to read my earlier posts above? How am I to keep up-to-date if my search is filtering out a relevant article such as “A point of view about stuttering” which a free search engine returns it as the very first hit? I need less of an explanation about DEVONagent’s limitations or affordances and more of a “this is how to do what you want to do using this program.” Thanks!


And of course disable all other filters (see Search Set window and Options tab of Search windows too)

The filter options are turned off. It’s still not returning many relevant articles, that otherwise appear at the very top of a Google Scholar search?

I’m trying to test these things you’re mentioning but now Google Scholar is returning 0 results no matter what I put into the search bar. For example, even Shakespeare returns 0 results.

Did you enter the suggested secondary query too? In addition, in case of too much traffic Google might temporarily block you.

Thanks for the second comment: it does appear that once I found the search in the log and proved I wasn’t a robot, results began appearing if I re-did the search.

What exactly does the secondary query do? I would assume that it might further limit rather than expand my search results, but that may be an inaccurate intuition. Is it adding multiple queries together rather than narrowing? And, to get the results I’m missing in the initial search require that I know what was filtered out before I do the secondary query? That’s a lot of extra work, since I’d have to find out what it missed and then help it not miss that relevant stuff(?).

It’s a wildcard and accepts every page. You could also enable the Express mode. This will deliver faster and unfiltered results. But especially the Digest is very limited or not available at all.

Thank you, I will play around with these different options this afternoon. I appreciate the help! (I’ve also come to determine that I may just need to make more use of the log when doing comprehensive, on demand searches (rather than scheduled crawls, etc.) It’s not ideal, but I also don’t think it’s the end of the world either. I’m able to access the exact same Google Scholar search result by accessing it from the log.)

You might also want to have a look at the PDF or eBook versions of the help (DEVONtechnologies | Handbooks and Extras). These formats are usually easier to read than the help in macOS’ help viewer.