Open Access News

News from the open access movement


Monday, April 17, 2006

Barbara Quint on Microsoft Academic Search

Barbara Quint, Microsoft Offers Alternative to Google Scholar: Windows Live Academic Search, Information Today Newsbreaks, April 17, 2006. Excerpt:

Microsoft has launched a test version of Windows Live Academic Search on its Live.com Web site...The new search tool will search proprietary content from scholarly publishers, as well as the open Web. The initial beta test offers content from 10 publishers; two more are on the way. Phase one of the beta concentrates on three subject areas: computer science, electrical engineering, and physics. A handsome interface offers users a polished approach to searching built around structured metadata supplied by publishers plus retrieval drawing on full-text spidering. Ironically, in the course of extensive interviewing, the two people who spoke most enthusiastically about the arrival of Microsoft into the academic/scientific “search space” were Anurag Acharya, the man behind Google Scholar, and Sharon Mombru, the woman running Scirus, Elsevier’s free sci-tech search engine. Although Microsoft representatives have been quoted as denying any “monetization” plans for the new service, I notice that the Web listings-in contrast with published sources-carry “Sponsored Links” advertising....

In contrast with Google Scholar’s resistance to publishing source information, Windows Live Academic Search has a full list of publishers and journal titles..., though dates of coverage aren’t indicated. Thiru Thirumalai-Anandanpillai (known to friends, colleagues, and, hereinafter, readers of this NewsBreak as “Thiru”), senior product manager at Microsoft, reported that the length of coverage varies from publisher to publisher. Some provide 10 or 15 years of content; a few offer 30 to 40 years. The network of joint publishing efforts, e.g., between commercial publishers and scholarly societies, makes the list of publishers very lengthy --it’s more than 100, in fact. The list of journal names has more than 7,100 titles, including 2,000-plus conferences. Overall, the system accesses some 8 million articles, according to Thiru....

Content alliances are being made in close association with CrossRef, the Publishers International Linking Association (PILA). CrossRef represents most major scholarly publishers, providing an infrastructure for linking citation metadata supplied by publishers through the OpenURL-compliant Digital Object Identifier (DOI) system. CrossRef also has a working arrangement with Google Scholar....Amy Brand, director of business development at CrossRef, saluted the new service: “We’re really pleased to be collaborating with Microsoft on an academic search tool that is designed both to improve the online research experience and to respect the concerns of the publishing industry. We look forward to helping bring even more content into Microsoft Academic Search as the initiative develops.” Brand described the relationship as “more of a ‘cooperation’ than a partnership per se.” The partnership includes “settling on standard terms and conditions for indexing publisher full-text.”

Also see Barbara's second article in the same issue, Windows Live Academic Search: The Details. Excerpt:

It will be months before we can expect a useful comparative review of Windows Live Academic Search and Google Scholar. Technically both products are in beta, but Google’s definition of the term “beta” covers some rather fully developed services, including Google Scholar (which is currently anticipating its second anniversary). Microsoft’s Academic Search is a true beta --it still adds significant content, and plans are already in place for a second beta phase....Nonetheless, the features and interface in Windows Live Academic Search have a very polished look. They offer users more sophisticated control of their searches than is usually seen in free services. Specifically, users can: [1] Sort and limit results by author, date (forward or reverse chronological order), journal, conference, or back to the default, relevance. [2] Use the “Richness Slider” to expand or contract the relevance ranked display of search results. [3] Click on author names for author bibliographies (with no limits on the numbers of authors “clickable”; no “et al.”). [4] View an abstract of each search citation in a preview pane section on the right of the screen or “hide abstract” if you just want to scan brief citations as quickly as possible using the full screen. [5] Export citation in basic text, RIS’ EndNote, or BibTeX bibliographic formats. [6] Find in a library using the connection to OCLC’s Open WorldCat service (on its way). [7] Save search strategies in macros using “Search Builder,” which will also support RSS feeds (“Feeds” at the top of the screen tied to a “+live.com” option) to alert searchers as new results appear on their Live.com page. (Can anyone here spell “SDI”?) Expect to see macro services in Academic Search within 2 weeks, according to Thiru Thirumalai-Anandanpillai (herein known as “Thiru”), senior product manager at Microsoft....If you want to restrict your search to more specific scholarly collections on the Web or to search on nontitle elements, one tip from Windows Live Academic Search staff is to enter your search statement along with the term “ArXiv” to reach content in the OAI-compliant institutional repositories which that source covers. By the way, Web content listings carry ads (“Sponsored Links”) while published sources do not....

Microsoft worked with CiteSeer in developing some of its features, e.g., author live links, and in tapping the Web-based content in which CiteSeer specializes, such as computer sciences. Thomson Scientific’s Institute for Scientific Information (ISI) also worked with CiteSeer in developing its Web Citation Index. However, Windows Live Academic Search does not offer forward or reverse citation indexing yet, as CiteSeer does. It does not automatically extract footnotes, though they may appear in the course of searching the full text. Again, this much-desired feature is on the planning boards, according to Thiru. Part of the problem involves standardizing author metadata (first name, last name, initials, punctuation, etc.), an area they are still working on, according to Thiru....Jill Grogg, electronic resources librarian at the University of Alabama Libraries, was “impressed and extraordinarily pleased that the initial release addressed OpenURL issues mmediately.” Grogg recalled that it took Google Scholar 4 to 5 months to support the linking and pointed out that Scirus still had not added it....She also commented on Microsoft’s using Open Archive for harvesting metadata: “It’s nice to see library-born standards being embraced by giants like Microsoft and Google.”...

Though the publishers have the option to provide Microsoft with the full abstract, first 140 characters from the abstract, or nothing, they must guarantee that nonsubscribers can at least see the full abstract as it appears on the publisher’s Web site. Publishers with PDF files or OAI-PMH standard repositories should have no problems, but if some technical difficulties occur, Microsoft promises to work through them....

Anurag Acharya, developer of Google Scholar, waxed lyrical about the arrival of a new player into his “search space.” “It’s entirely positive,” he said. “We all share the same problem, how to make all information easy to find. The more we can all do together, the better it is for the knowledge of the human race. It is fantastic that they are doing this.”