Google Scholar uses automated processes for indexing. To index a repository well, our search engine robots need to able to quickly reach all articles by following links from the home page and they need to be able to recognize bibliographic data for the article. If our robots are unable to find the URL for an article or are unable to fetch it, the article can not be included in the Scholar index. Furthermore, if they are unable to determine correct metadata for the article, we may not be able to identify citations to the article - which in turn will impact its ranking and visibility.
We encourage you to go over our inclusion guidelines at http://scholar.google.com/intl/en/scholar/inclusion.html to see if your repository meets the guidelines for indexing. We provide recommendations for some of the more common repository systems below. If you are using a repository system other than the ones mentioned, it would probably be good to discuss the guidelines with the provider of your repository software. We would be happy to discuss technical details as needed with your repository software provider.
Eprints: Repositories using Eprints version 3.0 or a later version are well-structured for indexing - search engine robots are able to quickly reach all articles by following links from the home page and individual article pages embed bibliographic data in a machine-readable fashion for indexing. If you are using a using an Eprints version earlier than 3.0 (May 2007), we strongly recommend upgrading to the latest version.
DSpace: Older versions of DSpace repositories have had some issues regarding indexing. We have worked with colleagues at DSpace to circumvent these problems. They have made several improvements which are included in recent releases. If you are using a DSpace release earlier than 1.7 (Dec 2010), we strongly recommend that you upgrade to the latest DSpace release. We also recommend that you go through the note that DuraSpace provides regarding indexing of DSpace repositories (https://wiki.duraspace.org/display/DSPACE/Ensuring+your+instance+is+indexed).
Digital Commons: Hosted Digital Commons repositories are well-structured for indexing - search engine robots are able to quickly reach all articles by following HTML links and individual article pages embed bibliographic data in a machine-readable fashion for indexing.