Monday, March 28, 2011

Alternative to Google Books: Hathi Trust signs Summon as search engine -


The Chronicle of Higher Education reports that HathiTrust has made its large library of digital texts searchable by signing an agreement with Summon, a library-specific search engine produced by Serials Solutions. Indeed, if you click on the HathiTrust link, you will see that they now offer catalog and full text searches, as well as browsing on their site.

Looking at the Search tips, they offer phrase searches in both the catalog and full text searches. They offer wild card searches, for both single and multiple characters, ONLY in the catalog search. They allow Boolean searching, with AND, OR, and NOT searches in both catalog and full text searches. You can also search subsets of books in full text searching. So you can select to search a private collection of books, for instance, in the full text search. There is advanced searching in the catalog search only, so far. They say they are "exploring" advanced search for the full text option. You cannot search non-text material, such as graphs or illustrations. There are much more details on searching, especially surrounding foreign language works.

If you are affiliated with a library that is part of the HathiTrust, you can get the entire text of a book that is in the public domain by searching HathiTrust. If you are not affiliated with such a library, what you get is a single page with your terms. If, however, the book is in the Google Books collection, you can go from the HathiTrust search on into GoogleBooks, and get it there. This information is provided at the HathiTrust Search Tips site under Printing/Downloading. They also remind users that they can locate many of the items in physical form in the libraries near them or by inter-library loan through those libraries. HathiTrust would like feedback from users who find problem pages, or otherwise have comments on the service.

The GoogleBooks Settlement which recently was rejected by the judge in the case, would have allowed the HathiTrust to post snippets of text along with search results, according to the Chronicle article. But the HathiTrust search results now will only display the page or pages of the document on which the search term appears. You can select the result, and retrieve the page or pages, and the term will be highlighted in a certain format. You can see either a "page view" which is like a PDF of the page, or a text view, where they will add the highlighting. (from HathiTrust SearchTips, under Search Tips. Readers who use screen readers and OCR software will want the text view, of course.

The decoration is the Hathi logo, an elephant, from their own website.

2 comments:

tom said...

Hello Betsy,

The full text search on the HathiTrust site is not powered by Summon. It uses software developed by HathiTrust on top of some open source software.

Tom Burton-West
Information Retrieval Programmer
University of Michigan Library
http://www.hathitrust.org/blogs/large-scale-search

Betsy McKenzie said...

Thank you! I am glad to have clarification from the elephant's mouth, as it were.