Search for unaccented versions of characters gives unexpected results

Search for thomas a kempis (without the accent over the à, since it's hard to type in). The top two results you'd expect to be found (The Imitation of Christ and The Works of Thomas à Kempis) end up as positions 8 and 9 in the search results, and it's only incidental that they were found at all (because the letter a without accent happens to appear elsewhere in the page; it is highlighted in the sample shown with the search hits, whereas the à in thomas à kempis is not highlighted). If you search for thomas kempis or thomas à kempis these two resources will be hits number 1 and 2.

You should ignore the accent on accented characters (or rather, match both unaccented and accented forms if the accent is missing in the search text), as Google does. It ought to be possible to search for something without having to type in all its accents.

Find more posts tagged with

Comments

Keep Smiling 4 Jesus :)

If you search for thomas kempis or thomas à kempis these two resources will be hits number 1 and 2.

A search for kempis has same results for number 1 and 2 since kempis is unique among author names.

By the way, Logos 4 library search ignores accent so have same results.

Keep Smiling [:)]

fgh

I would also appreciate if Logos could teach its website that an encyclopaedia is the same thing as an encyclopedia, and that archaeology is the same thing as archeology. Currently a search for catholic encyclopaedia does not find the Catholic Encyclopedia, and a search for archaeology results in a completely different list of books than a search for archeology. I shouldn't have to know which spelling a book uses in order to be able to search for it... Especially since there are sometimes both British and American editions of the same book, and Logos sometimes uses the one and sometimes the other, depending on which gets out of copyright first or which publisher is easier to work with.

Graham Criddle

https://community.logos.com/discussion/comment/315779#Comment_315779

[Y][Y]

Rich DeRuiter

You should ignore the accent on accented characters (or rather, match both unaccented and accented forms if the accent is missing in the search text), as Google does. It ought to be possible to search for something without having to type in all its accents.

I expect the search engine Logos uses for its sites are not developed internally. That means they couldn't fix what you're asking them to fix - at least not that way. The same for FGH's issue. I suspect the same might be true where British vs. American spelling is concerned (favour vs. favor. e.g.).

What might work, though is if Logos web dev's could identify these issues and include generic and alternative spellings in the part of the metadata their search engine would find, or in the displayed text (hidden? at the bottom? white on white?). That would be a massive undertaking, but might be easier than amending their site search engine (if that's even possible), or developing their own search engine.

Rosie Perera

https://community.logos.com/discussion/comment/315788#Comment_315788

I expect the search engine Logos uses for its sites are not developed internally.

....

That would be a massive undertaking, but might be easier than amending their site search engine (if that's even possible), or developing their own search engine.

What? Logos is essentially a search engine company. That's their primary expertise in their Bible software. You think they didn't develop their own website's search engine, with all its cool faceting and stuff? That would be mind boggling, though I suppose it's conceivable.

And even if they did buy the code from somewhere else (as they did with the forum software), I'm sure do have the ability to tweak it (as they've done with the forum software).