Date: Tue, 8 Apr 86 09:01:47 est From: vtisr1!irlistrq To: fox Subject: IRList Digest V2 #18 Status: RO IRList Digest Tuesday, 8 Apr 1986 Volume 2 : Issue 18 Today's Topics: Query - Address for UW Centre for the new OED Discussion - Re "Keyword Consistency and Thoroughness" by J.H. Coombs Discussion - Re Intelligent Front Ends, CONIT Announcement - Schedule for next Forum issue Announcement - Faculty Position Opening in Information Studies ---------------------------------------------------------------------- Date: 1986 Mar 28 10:52 SET From: Robert Philip Weber Subject: UW CENTRE FOR THE NEW OXFORD ENGLISH DICT NEWSLETER In V2 no 11 you reprint some info from the UW New Oxford Dict Newsletter. Do you have either snail mail or email addresses so that I can follow up on some of the info? Thanks. Bob Weber BITNET : Weber3@Harvarda ARPA: Weber3%Harvarda.Bitnet@Wiscvm.wisc.edu (i am in sweden at the moment, but my email is forwarded automatically from harvard. Thanks.) [Note: Address for newsletter is Univ. of Waterloo, Waterloo, Ontario, Canada N@L 3G1 (519) 885-1211 x6183. In No. 6, Sept. 1985 it mentions that Maureen Searle is serving as secretary, in case you want a person to send to. Does anyone have an electronic address? - Ed] ------------------------------ From: Aviezri Fraenkel Date: Sun, 30 Mar 86 12:01:24 -0200 Subject: Keywords... Regarding "Keyword Consistency and Thoroughness" by J.H. Coombs (IRList #17), the establishment of good heuristics for the selection of keywords is indeed a problem which needs to be addressed more fundamentally than has been done hitherto, to my knowlege. THE PROBLEM IS OLD. Let me quote from Colin Tapper, British experience in legal information retrieval, MULL pp. 127-134 (Dec 1964): "The plan...was to index all the material in a given area for both legal concepts and fact situations, and then to attempt sample searches. The area selected was the law relating to the admissibility in evidence of confessions. It was felt that this area was well-defined, self-contained, and, in this country at least, of relatively small compass. "Some preliminary indexing was carried out on this basis. It proved to be disastrous. For one thing, the legal concepts, so well-defined in books, proved to be much less so in practice. It was found that the indexers were quite inconsistent in their attempts to index the same document, and that even the same indexer was liable to index a document differently at different times. This was especially true when it was presented in a different context." THE PROBLEM IS DIFFICULT. Modelling the keyword selection process, it has been shown that the easy part of it, namely the question whether there is a bounded number of keywords for representing a document {the bounded thoroughness problem}, is NP-complete (Proc. 4th Intern. Conf. on Inf. Storage & Retrieval, Oakland, CA, 1981, pp. 77-82, ACM 1981). These remarks of course do not solve the indexing predicaments, but perhaps they contribute ever so little to convince IR researchers -- those that still need convincing! -- that the problems are difficult, and new and more powerful tools for their solution are required... Regards, Aviezri S. Fraenkel. ------------------------------ Date: Sun, 6 Apr 86 22:12 EST From: "Richard (Dick) Marcus" Subject: CONIT Intermediary = Intelligent Front End cc: Gollu@MIT-MULTICS.ARPA Message-ID: <860407031256.924813@MIT-MULTICS.ARPA> Thanks to Alan Smeaton and Editor Ed for referencing [IRD:2(17)] our CONIT retrieval assistant system in regard to Hans Peter Giger's request [IRD:2(15)] for info on intelligent front ends. CONIT is an experimental syystem for investigating assisting users retrieve information, especially bibliographic or text-based, from heterogeneous systems and databases. The last major journal article on CONIT was in Journal of the ASIS (JASIS) 1983: 34(6):381-404. That full reference and abstract appear as entry 3 in the bibliography prepared by Salton and Raghavan in IRD:2(7). Other papers, articles and references appear in JASIS (1981), IEEE Transactions on Systems, Man, and Cybernetics (SMC-1982), Annual Meeting Proceedings of ASIS in 1980-83 and 85, ACM SIGIR FORUM 1974, RIAO Conference (Grenoble, France. 1985), and ACM/BCS Symposium (Poster Presentation -- Cambridge, England. 1984). Work in this area has been accelerating in recent years as attested to by the fact that we are conducting a SECOND conference on the subject next month (see item FOLLOWING Smeaton item in IRD:2(17)). The fact that Smeaton couldn't find a reference to CONIT may say something about the inadequacies of our existing IR systems. In this respect, we have a problem of terminology as pointed out by James Coombs in the PRECEDING item of IRD:2(17): front ends are also called interfaces, intermediaries, gateways, assistants, etc. Also, calling something intelligent, or expert, or Knowledge-based doesn't make it so and, conversely, not calling it intelligent doesn't mean it isn't. None of this will be news to Giger with whom I have been communicating and who is getting CONIT demos via networking to the MIT computer by a student of mine, Aleks Gollu, who is spending the Spring term at ETH in Zurich. ---Dick Marcus... ------------------------------ Subject: msg indicating Forum schedule Date: Sun, 30 Mar 1986 23:36 CST From: Vijay V. Raghavan Dear Ed, I received a copy of your message to Mike Lesk in which you ask about the status of Forum. I have been busy and it hasn't gone out yet. I expect it will be out in a week or so. Certainly, it will be mailed to ACM before April 10th. With best wishes, Vijay [Note: People can begin sending submissions (to Vijay) for next issue. This one should be mailed before the end of May. - Ed] ------------------------------ Date: 31 March 1986, 22:43:22 LCL From: RNODDY at SUVM.bitnet Subject: faculty position at Syracuse University ... the IRlist Digest. Very interesting! Vijay accepted my ad for the next issue of SIGIR Forum. In case you are curious, here is a copy: NEW TENURE-TRACK FACULTY POSITION ASSISTANT PROFESSOR The Syracuse University School of Information Studies is seeking an individual for a full-time, tenure-track faculty position beginning September, 1986. The teaching and research interests of the School include the following areas, in the context of their application in information transfer: Natural Language Processing, Human-Computer Interaction, Telecommunications/Networking, Database Management. Preference will be given to those applicants who best reflect one or more of these areas and complement the other needs of the School We encourage applications from individuals with a variety of backgrounds, such as information science, behavioral sciences, computer science, linguistics, and library science. Applicants should have a commitment to excellence in teaching and a proven research/publication record commensurate with experience. PhD is required. The closing date for applications is May 15, 1986, or until filled. Applications should be addressed to: Dr. Robert N. Oddy, Chair, Search Committee, School of Information Studies, 200X Huntington Hall, Syracuse University, Syracuse NY 13244-2340. (On Bitnet: RNODDY at SUVM). Syracuse University is an AA/EOE. Yours, Bob Oddy. ------------------------------ END OF IRList Digest ********************