Date: Mon, 8 Mar 1993 10:29:50 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: Charlotte Skuster Subject: Ooops! Will the person who sent the message about abstracting for Compendex/Ei please send the message again? I discarded it by mistake. Sorry. Charlotte ========================================================================= Date: Mon, 8 Mar 1993 10:33:43 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: Jian Liu Subject: Re: Checklist of Library Software ----------------------------Original message---------------------------- I got it two days ago without any problem. I just printed it today. Tons of information in a very presentable format. On Fri, 5 Mar 1993 13:52:47 ECT, R.S. Etheredge wrote: >----------------------------Original message---------------------------- >Howdee, >Regarding the message on getting via ftp a document on library >software from Erik Delfino, the ftp server seq1.loc.gov >(140.147.3.12) seems to be unavailable. I time out, so a connection >cannot be established. Any body else have this problem? Generally, I >have been able to ftp anywhere I want to... >Thanks for your attention, and have a happy day... >Rusty Etheredge _____________________ Liu Jian Indiana University jiliu@ucs.indiana.edu --------------------- ========================================================================= Date: Mon, 8 Mar 1993 11:32:37 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: "Linda Musser -- Penn State (lrm@psulias.bitnet)" Subject: Forwarded message re student abstractors ----------------------------Original message---------------------------- The following message was originally posted to ELDNET-L, a listserv for engineering librarians. I thought it might be of some interest to this group. I have edited out some of the extraneous text but left in the parts relating to the program of student abstractors. Linda Musser Penn State ================================FORWARDED NOTE=========================== Date: Fri, 05 Mar 1993 10:41:38 CST From: shimek@kirk.msoe.edu Subject: A Reply to Concerns About Engineering Index/Compendex coverage ... I am the coordinator of the MSOE-Engineering Information Abstracting Project here at the Milwaukee School of Engineering. The project, which began more than two years ago, is a joint agreement between MSOE and Engineering Information Inc. (Ei), producers of the Engineering Index/Compendex. In this agreement, MSOE essentially contracted to become abstract suppliers for Ei. As coordinator of the project, among other responsibilities, I hire and train academically- qualified, upperclass engineering students to read, to evaluate critically and to abstract articles from publications representing the entire coverage of Engineering Index and Compendex. Students are often advised on the technical content of articles by expert faculty and staff. The project is extremely demanding, but it has also elicited highly favorable responses from ABET, potential employers and others interested in an engineering education which offers students the opportunity to develop solid writing skills and to study the current professional literature from all fields of engineering. To date, we have produced almost 15,000 abstracts for Ei, which amounts to a very modest percentage of Ei's annual output. Among those 15,000, however, are a number of abstracts from various issues of _Civil Engineering_. Last month, for example, one of my students completed abstracts for the following _Civil Engineering_ issues: September 1991, January 1992, May 1992 and June 1992. Undoubtedly, a real problem highlighted here, however, is the notion of currency. Ei does indeed index and abstract _Civil Engineering_, but in this case, there is an unusual lag-time between publication of particular issues of the journal and their appearance in Engineering Index/Compendex. In the fields of engineering, where currency obviously is highly valued, this delay is regrettable. I have contacted Ei about the delay. The problem concerned something often faced by libraries: catching up with subscription claims. It is NOT the case that Ei is not abstracting _Civil Engineering_, or other ASCE publications. Based on what we index and abstract here, I can also confirm that this type of delay is rare with Ei. The vast majority of documents with which we work generally have been published within the past few months, if not the past few weeks. Moreover, Ei holds us to very strict publishing deadlines. In general, then, not only do we work with current material (at least, as current as possible with print stuff), but we also turn it around within a week. I hope that this explanation clarifies concerns regarding the coverage of Engineering Index/Compendex. More importantly, however, I would like to take this opportunity to solicit other comments, suggestions, feedback, etc. from librarians and information professionals regarding Engineering Index/Compendex, and any additional Ei products, such as the new thesaurus. A major flaw in the new thesaurus, of course, is the lack of CAL code mapping, which will be corrected in the next edition. I'd be happy to pass any comments along to Ei. In addition, as an engineering librarian who writes and edits technical abstracts for Ei, I'd welcome any comments which would help me and participants in our project to produce better abstracts for our customers. Thankyou, Gary Shimek Assistant Director INTERNET: SHIMEK@KIRK.MSOE.EDU Ei Abstracting Project Coordinator Walter Schroeder Library Milwaukee School of Engineering ========END OF FORWARDED NOTE=========== ========================================================================= Date: Tue, 9 Mar 1993 14:19:13 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: David Lewis Subject: Query: Length of typical queries to online retrieval systems ----------------------------Original message---------------------------- Hi -- An employee of one of the big online database companies once told me that 90% of the queries submitted to their system were one word long, despite their provision of a rich searching language. My own experience with demoing natural language interfaces to statistical text retrieval systems has also suggested that people typically enter very short queries, even though longer queries would be more effective. However, I don't know of any published figures on the distribution of lengths of queries made to operational systems, degree of use of AND/OR/etc., and so on. Can anyone provider pointers to such articles? If you reply to me I will post a summary of replies to this list, and will include your name along with your comment unless otherwise requested. thanks, Dave David D. Lewis AT&T Bell Laboratories email: lewis@research.att.com 600 Mountain Ave.; Room 2C-408 ph. 908-582-3976 Murray Hill, NJ 07974; USA dept. fax. 908-582-7550 ========================================================================= Date: Wed, 10 Mar 1993 10:37:49 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: Jessica Milstead <76440.2356@CompuServe.COM> Subject: Standard for arrangement ----------------------------Original message---------------------------- The NISO committee which is revising the standard for indexes has recommended that NISO develop a standard for arrangement (aka filing). As the person covering this area on NISO's Standards Development Committee, making a recommendation on development of a standard is my responsibility. The issue is not a simple one, and I am asking for input from interested parties. Some background: In the international arena, ISO has a standard set of "principles" for arrangement, and a technical report which consists of an illustrative set of rules. This standard is a worthy effort, but it is very difficult to read and understand -- a severe limit on its usefulness. The British standard for arrangement is a set of rules. Jim Anderson has given me permission to quote from a forthcoming paper of his which states the U.S. situation very eloquently. He shows the pitfalls which NISO must avoid if it does decide develop a standard. "In the 1970s, both the American Library Association and the Library of Congress got tired of filing numeals as if they were spelled out in hundreds (if not thousands) of different languages (or perhaps it was their computers that got tired). And they got tired of having to figure out what abbreviations stood for. So they decided they needed new filing rules. Did they commission research? Did they seek empirical data on what arrangements real users found most easy to comprehend and use? No, they convened committees of experts, who drew up rules based on their expert opinion; they circulated drafts and considered opinions, and finally settled on what they thought was best. But they, LC and ALA, disagreed sharply on some key points, so we now have two quite different de facto standards for library filing in the United States. "In case you are wondering where the disagreement is, ALA says that you can't expect users to understand any basis for arranging headings other than the actual alphanumeric characters in the headings, so the ALA rules call for strictly (almost) alphanumeric order. "LC still insists on structuring the arrangement for headings that begin witht he same word, so that you have separate little (and sometimes not so little alphabetical arrays of forenames, then surnames, then places, then other things, then titles, so that a title like "Charles and the Wisemen" will file long after "Charles, Virginia," representingeither Virginia Charles, a person, or a place in Virginia. There are even more little arrays (or not so little) among subject headings, because LC insists on arranging different types of subheadings by criteria other than spelling. "Disagreement on these points in the indexing community is even more marked, with the Chicago Style Manual still suggesting that letter by letter ifling might be OK, while the entire library community, and the majority of our committee, have decided (on the basis of our wisdom, rather than any research) that at least, headings that begin with the same word should file together, and therefore spaces must be considered in filing." So, I have a set of questions: Should NISO try to develop a single standard for the U.S.? If so, should it be a set of principles, with one or more examples, or should we attempt to legislate one set of practices? How do we justify whatever we actually do -- and persuade people to accept it? Please let me have your comments. I should alert you now, however, that NISO is probably no more likely than anyone else to commission research on the issue -- problems of both funding and design inhibit us. ========================================================================= Date: Wed, 10 Mar 1993 14:03:20 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: Alicia Abramson Subject: Spreading Activation ----------------------------Original message---------------------------- I am a student doing some research on spreading activation and its use/potential applications in indexing. I am particularly interested in finding out if anyone knows of any software that actually uses algorithms for spreading activation. In addition, any references to articles/studies about spreading activation and how and if it models cognitive processes would also be of help. Finally, I am interested in finding out if actual "real - life" indexers have used, or come across spreading activation in their own work. Alicia Abramson alicia@info.berkeley.edu MLIS Student or UC Berkeley l200-dn@garnet.berkeley.edu School of Library and Information Studies ========================================================================= Date: Wed, 10 Mar 1993 16:39:21 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: KROVETZ@cs.umass.EDU Subject: Re: spreading activation ----------------------------Original message---------------------------- I don't know of an papers about spreading activation for indexing, but the following paper discusses spreading activation for retrieval: "Information Retrieval by Constrained Spreading Activation in Semantic Networks", P. Cohen and R. Kjeldsen, Information Processing and Management, Vol. 23(4), pp. 255-268, 1987 Bob krovetz@cs.umass.edu ========================================================================= Date: Thu, 11 Mar 1993 10:06:30 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: Mary S Stephenson Subject: Software to produce a "Hansard's" ----------------------------Original message---------------------------- Has anyone had any experience with database or indexing software used to produce a "Hansard's" for parliamentary/legislative debates? It needs to be capable of using a record format that will all the use of subject access, but does not need to support full-text. It also should support the use of Boolean retrieval as well as (hopefully) proximity and truncation. Finally, it will need to be used to produce a printed annual version of the database. Please reply to me rather than the list. Any and all help would be much appreciated. Mary Sue Stephenson Univ. of British Columbia SLAIS mss@unixg.ubc.ca 604-822-6392- ========================================================================= Date: Thu, 11 Mar 1993 10:06:59 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group From: "Kate McCain" Subject: Re: Spreading Activation In-Reply-To: Message of Wed, 10 Mar 1993 14:03:20 ECT from ----------------------------Original message---------------------------- An interesting question. I have actually used examples of "spreading activation theory" (and concept maps a la AI) in teaching my A&I class. The only "example" that I can think of in application, other than "thinking about thinking" is a paper that was in Information Processing & Management a few years ago that used the SAT-type conceptual networks as a foundation for building a grant proposal information retrieval system. I think it was in the special issue edited by Bruce Croft; I'm sorry that I don't have the reference to hand. The issue of links between concepts and the attenuation of linkages as you get further away from the stimulus is also likely to play a part in at least some neural network applications that may be indexing/information retrieval relevant. I would try some citation searching based on the original SAT work (Loftus??). After all -- the idea of "training" neural networks is the idea of strengthening some patterns of association at the expense of others. Good luck. Hope this helps. Sounds very interesting. Kate McCain "bibliometrics R us" College of Information studies mccainkw@duvm.ocs.drexel.edu ========================================================================= Date: Thu, 11 Mar 1993 16:45:21 ECT Reply-To: Indexer's Discussion Group Sender: Indexer's Discussion Group Comments: Resent-From: Charlotte Skuster From: Charlotte Skuster ----------------------------Original message---------------------------- Received: from BINGVMB.BITNET by BINGVMB (Mailer R2.08 PTF008) with BSMTP id 6059; Thu, 11 Mar 93 10:16:11 ECT §MSG:FROM: JBRINKMA--ESRIN TO: INDEX-L --BINGVMB 93-03-11 16:16:50 To: INDEX-L --BINGVMB *** Reply to note of 93-03-11 16:10 From: Jochen Brinkmann IRS Subject: Software to produce a "Hansard's" Please reply to the list. It's an interesting subject. Thanks End of Message ÿÿ $ Software to produce a "Hansard's" -R =========================================================================