From FAQ comp.ai.shells found at ftp:///rtfm.mit.edu:/pub/usenet-by-group/comp.ai.shells Status: RO ------------------------------------------------------------------- Subject: [1-5a] Free/Cheap Expert System Shells Remember, when ftping compressed or compacted files (.Z, .gz, .arc, .fit, etc.) to use binary mode for retrieving the files. Files that end with a .gz suffix were compressed with the patent-free gzip (no relation to zip). Source for gzip is available from: prep.ai.mit.edu:pub/gnu/{gzip-1.2.3.shar,gzip-1.2.3.tar,gzip-1.2.3.msdos.exe} If you do not have ftp access, you can FTP files by E-mail. Send a message with the word "help" in the body to ftpmail@decwrl.dec.com. FOCL is an expert system shell and machine learning program written in Common Lisp. The machine learning program extends Quinlan's FOIL program by containing a compatible explanation-based learning component. FOCL learns Horn Clause programs from examples and (optionally) background knowledge. The expert system includes a backward-chaining rule interpreter and a graphical interface to the rule and fact base. For details on FOCL, see: Pazzani, M. and Kibler, D., "The role of prior knowledge in inductive learning", Machine Learning 9:54-97, 1992. It is available by anonymous ftp from ics.uci.edu:/pub/machine-learning-programs/ as the files README.FOCL-1-2-3, FOCL-1-2-3.cpt.hqx (a binhexed, compacted Macintosh application), FOCL-1-2-3.tar.Z (Common Lisp source code), and FOCL-1-2-3-manual.hqx (binhexed manual). If you use a copy of FOCL, or have any comments or questions, send mail to pazzani@ics.uci.edu. SOAR -- ftp.cs.cmu.edu: /afs/cs.cmu.edu/project/soar/public/Soar5/ -- Lisp Version /afs/cs.cmu.edu/project/soar/public/Soar6/ -- C Version Contact: soar-request@cs.cmu.edu Integrated Agent Architecture. Supports learning through chunking. OPS5 -- ftp.cs.cmu.edu:/user/ai/areas/expert/systems/ops5/ops5.tar.gz BABYLON is a development environment for expert systems. It includes frames, constraints, a prolog-like logic formalism, and a description language for diagnostic applications. It is implemented in Common Lisp and has been ported to a wide range of hardware platforms. Available by anonymous ftp from ftp.gmd.de:/gmd/ai-research/Software/Babylon/ [129.26.8.84] as a BinHexed stuffit archive, on the Web via the URL http://www.gmd.de/ on the Apple CD-ROM, or with the book "The AI Workbench BABYLON", which contains *full source code* of BABYLON and the stand-alone version for the Mac. The book describes the use of BABYLON in detail. MOBAL is a system for developing operational models of application domains in a first order logic representation. It integrates a manual knowledge acquisition and inspection environment, an inference engine, machine learning methods for automated knowledge acquisition, and a knowledge revision tool. By using MOBAL's knowledge acquisition environment, you can incrementally develop a model of your domain in terms of logical facts and rules. You can inspect the knowledge you have entered in text or graphics windows, augment the knowledge, or change it at any time. The built-in inference engine can immediately execute the rules you have entered to show you the consequences of your inputs, or answer queries about the current knowledge. MOBAL also builds a dynamic sort taxonomy from your inputs. If you wish, you can use several machine learning methods to automatically discover additional rules based on the facts that you have entered, or to form new concepts. If there are contradictions in the knowledge base due to incorrect rules or facts, there is a knowledge revision tool to help you locate the problem and fix it. MOBAL (release 3.0b) is available free for non-commercial academic use by anonymous ftp from ftp.gmd.de:/gmd/mlt/Mobal/ The system runs on Sun SparcStations, SunOS 4.1, and includes a graphical interface implemented using Tcl/TK. MIKE (Micro Interpreter for Knowledge Engineering) is a full-featured, free, and portable software environment designed for teaching purposes at the UK's Open University. It includes forward and backward chaining rules with user-definable conflict resolution strategies, and a frame representation language with inheritance and 'demons' (code triggered by frame access or change), plus user-settable inheritance strategies. Automatic 'how' explanations (proof histories) are provided for rule exectuion, as are user-specified 'why' explanations. Coarse-grained and fine-grained rule tracing facilities are provided, along with a novel 'rule graph' display which concisely shows the history of rule execution. MIKE, which forms the kernel of an Open University course on Knowledge Engineering, is written in a conservative and portable subset of Edinburgh-syntax Prolog, and is distributed as non-copy-protected source code. MIKE version 1 was described in the October/November 1990 issue of BYTE. MIKE v1.50, which was formerly available from a range of ftp servers, has been superseded by two newer versions: MIKEv2.03, a full Prolog source code version, incorporating a RETE algorithm for fast forward chaining, a truth maintenance system, uncertainty handling, and hypothetical worlds, and MIKEv2.50, a turnkey DOS version with menu-driven interface and frame- and rule-browsing tools, fully compatible with MIKEv2.03, but without source code. They are available by anonymous ftp from hcrl.open.ac.uk [137.108.81.16] as the files MIKEv2.03: /pub/software/src/MIKEv2.03/* MIKEv2.50: /pub/software/pc/MIKEV25.ZIP They are also available from the CMU AI Repository. For further information, please contact Marc Eisenstadt, M.Eisenstadt@open.ac.uk, Human Cognition Research Lab, The Open University, Milton Keynes MK7 6AA, UK, phone +44 908-65-3149, fax +44 908-65-3169. ES: The October/November 1990 issue of BYTE also described the ES expert system. ES supports backward/forward chaining, fuzzy set relations, and explanation, and is a standalone executable for IBM-PCs. ES is available by anonymous ftp from ftp.uu.net:/pub/ai/expert-sys/ [192.48.96.9] as summers.tar.Z. ftp.uu.net is mirrored on unix.hensa.ac.uk [129.12.21.7] under /pub/uunet/. GEST (Generic Expert System Tool) Version 4.0 is a flexible expert system shell applicable to a variety of problem domains. Its knowledge representation schemes include facts, frames, rules, and procedures. It supports forward and backward inferencing with multiple options for each and message handling for frame objects. There is a time system to support an event history and reverting back to a previous state. In addition, there is support for fuzzy logic and certainty factor maintenance. Features new to V4.0 include a blackboard architecture (single level and hierarchical) with a control module and knowledge sources. The user interface utilizes the Symbolics windowing system and is menu and mouse driven. GEST can be embedded in a user program (with or without the GEST user interface) by using a library of function calls. The printed reference manual includes a tutorial, table of contents, and thorough index. GEST runs only on Symbolics Lisp Machines, Genera 7.2. GEST users are expected to be familiar with Symbolics machines. Source and object files are available by anonymous ftp to ftp.gatech.edu in the (binary) file pub/ai/gest.tar.Z. Register your copy with John Gilmore, Georgia Tech Research Institute, ITL/CSITD 0800, Atlanta, GA 30332, USA, Internet: john.gilmore@gtri.gatech.edu. Send your name, address, organization, phone number, and Internet email address. To request a free printed copy of the GEST manual (not in the distribution file), please also contact John Gilmore. Users are free to produce as many copies of the GEST Reference Manual as they require without any restrictions. WindExS (Windows Expert System) is a fully functional Windows-based forward chaining expert system. Its modular architecture allows the user to substitute new modules as required to enhance the capabilites of the system. WindExS sports Natural Language Rule Processor, Inference Engine, File Manager, User Interface, Message Manager and Knowledge Base modules. It supports forward chaining, and graphical knowledge base representation. Write etoupin@aol.com for documentation and operational system. RT-Expert is a shareware expert system that lets C programmers integrate expert systems rules into their C or C++ applications. RT-Expert consists of a rule-compiler that compiles rules into C code, and a library containing the rule execution engine. RT-Expert for DOS works with Borland Turbo C, Borland C++, and Microsoft C/C++ compilers. The personal edition is licensed for educational, research, and hobby use. Applications created with RT-Expert personal edition are not licensed for commercial purposes. Professional editions are available for commercial applications using DOS, Windows, and Unix environments. RT-Expert is available by anonymous ftp from world.std.com:/vendors/rtis/rtexpert For more information, write to Real-Time Intelligent Systems Corporation . ---------------------------------------------------------------- Subject: [1-5b] Free/Cheap Expert System Shells: CLIPS and Related Systems CLIPS 6.0 (C Language Integrated Production System) is an OPS-like forward chaining production system written in ANSI C by NASA. The CLIPS inference engine includes truth maintenance, dynamic rule addition, and customizable conflict resolution strategies. CLIPS, including the runtime version, is easily embeddable in other applications. CLIPS includes an object-oriented language called COOL (CLIPS Object-Oriented Language) which is directly integrated with the inference engine. CLIPS runs on many platforms including IBM PC compatibles (including Windows 3.1 and MS-DOS 386 versions), Macintosh, VAX 11/780, Sun 3/260, and HP9000/500. CLIPS is available from COSMIC at a "nominal" fee (the MS-DOS/Windows 3.1 version $350 for the software and $115 more for the documentation, with discounts for US academic institutions; for update orders $100 and $200, respectively) for unlimited copies with no royalties. (CLIPS is available free to NASA, USAF, and their contractors for use on NASA and USAF projects.) For more information, send email to service@cossack.cosmic.uga.edu, write COSMIC, University of Georgia, 382 East Broad Street, Athens, GA 30602, call 706-542-3265, or fax 706-542-4807. To subscribe to the CLIPS mailing list, send a message to the list server listserv@cossack.cosmic.uga.edu (128.192.14.4) with message body SUBSCRIBE CLIPS-LIST The CLIPS help desk phone number is 713-286-8919 (fax 713-286-4479/244-5698) and email address is stbprod@fdr.jsc.nasa.gov. (The address is STB Products Help Desk, LinCom Corporation, 1020 Bay Area Boulevard, #200, Houston, TX 77058-2628.) The Software Technology Branch's home page is http://www.jsc.nasa.gov/stb/STB_homepage.html NASA Information Services's home page is http://hypatia.gsfc.nasa.gov/NASA_homepage.html and the CLIPS home page is http://www.jsc.nasa.gov/~clips/CLIPS.html [Note: Folks who obtain CLIPS from COSMIC can do anything they wish with it, including redistribute it. Folks who obtain CLIPS directly from NASA are restricted to using it for US government purposes only.] The CLIPS FAQ list and bug fixes are available from hubble.jsc.nasa.gov:/pub/clips/ and are maintained by Gary Riley . Joseph Giarratano and Gary Riley's book, "Expert Systems Principles and Practice", comes with an MS-DOS CLIPS 6.0 interpreter (see [1-3] above). CLIPS is also available on the various AI CD-ROMs, and the MS-DOS version is available for $50 from Austin Code Works , 11100 Leafwood Lane, Austin, Texas 78750-3409, phone 512-258-0785, fax 512-258-8831, or BBS 512-258-8831. DYNACLIPS (DYNAamic CLIPS Utilities), is a set of blackboard, dynamic knowledge exchange, and agent tools for CLIPS 5.1 and 6.0. It is implemented as a set of libraries that can be linked with CLIPS 5.1 or CLIPS 6.0. Versions 3.0 and 3.1 will work with either CLIPS 5.1 or CLIPS 6.0. Source code is not available. Agents use the blackboard to communicate with other intelligent agents in the framework. Each intelligent agent can send and receive facts, rules, and commands. Rules and facts are inserted and deleted dynamicly while the agents are running. Knowledge can be transfered on a temporary or permanent basis. For more information, please contact Yilmaz Cengeloglu, PO Box 4142, Winter Park, FL 32793-4142, or send email to cengelog@escmail.orl.mmc.com, yil@engr.ucf.edu or 73313.775@compuserve.com. It is available from the CMU AI Repository in ftp.cs.cmu.edu:/user/ai/areas/expert/systems/clips/dyna/ FuzzyCLIPS 6.02 is a version of the CLIPS rule-based expert system shell with extensions for representing and manipulating fuzzy facts and rules. In addition to the CLIPS functionality, FuzzyCLIPS can deal with exact, fuzzy (or inexact), and combined reasoning, allowing fuzzy and normal terms to be freely mixed in the rules and facts of an expert system. The system uses two basic inexact concepts, fuzziness and uncertainty. Versions are available for UNIX systems, Macintosh systems and PC systems. There is no cost for the software, but please read the terms for use in the FuzzyCLIPS documentation. FuzzyCLIPS is available via WWW (World Wide Web). It can be accessed indirectly through the Knowledge Systems Lab Server using the URL http://ai.iit.nrc.ca/home_page.html or more directly by using the URL http://ai.iit.nrc.ca/fuzzy/fuzzy.html or by anonymous ftp from ai.iit.nrc.ca:/pub/fzclips/ For more information about FuzzyCLIPS send mail to fzclips@ai.iit.nrc.ca. wxCLIPS provides a simple graphical front end to CLIPS 5.1, and in addition contains a suite of extra functions for creating GUI applications in CLIPS. wxCLIPS is available as Sun Open Look, Sun Motif, Linux Open Look, and Windows 3.1 binaries. wxCLIPS is available by anonymous FTP from skye.aiai.ed.ac.uk:/pub/wxclips/ [192.41.104.6] or the URL http://www.aiai.ed.ac.uk/~jacs/wxclips/wxclips.html To be added to the wxclips-users or wxclips-announce mailing lists, send mail to wxclips-users-request@aiai.edinburgh.ac.uk. For more information, write to Julian Smart . ----------------------------------------------------------------