We are releasing a slightly extended version 6.7 of Attempto Controlled English (ACE) and an improved version 6.7 of the Attempto Parsing Engine (APE).

Changes in ACE 6.7

• Added support for the exponentiation operator ^ in arithmetical expressions, e.g. "2^3^4", which is parsed as (2^3)^4.
• Added support for how many and how much queries, which ask for numbers of countable nouns or for amounts of mass or measurement nouns, e.g. "How many beaches have how much sand?"

Changes in APE 6.7

• APE can now translate ACE texts into the RuleML format.
• The coverage of the DRS-to-OWL translator has been improved.
• The dependencies of the Java Interface for APE on external libraries are now managed using Maven.
• The large Clex lexicon (~100,000 entries) was migrated to a separate repository Attempto/Clex. As before, the APE source distribution contains a reduced (~2000 entries) version of this lexicon.

The source code of APE can be downloaded from

https://github.com/Attempto/APE/releases

The large Clex lexicon can be downloaded from

https://github.com/Attempto/Clex/releases

Detailed documentation on ACE 6.7 and its tools is available on the Attempto website that was updated and partially restructured.

Please use the Attempto mailing list for any feedback. Technical contributions to APE can also be made via joining the APE project on GitHub.

During the last two years, the Attempto group was involved in the EU project MOLTO. Our involvement resulted in two new tools.

• ACE-in-GF is a multilingual grammar for ACE, i.e. tool that bidirectionally maps ACE sentences to fragments of several natural languages.
• AceWiki-GF is a multilingual extension of AceWiki.

For related publications, demos and source code, see the website of each tool.

Program Announcement and Call for Participation: Third Workshop on Controlled Natural Language (CNL 2012)

29-31 August 2012
Zurich, Switzerland

http://attempto.ifi.uzh.ch/site/cnl2012/

We are pleased to invite you to participate at the Third Workshop on Controlled Natural Language (CNL 2012).

The program of CNL 2012 consists of two invited talks and twelve research presentations:

http://attempto.ifi.uzh.ch/site/cnl2012/program.html

In addition, there will be two demo slots for industrial and academic demos:

http://attempto.ifi.uzh.ch/site/cnl2012/demos.html

There are still free demo slots available. Let us know if you would like to demonstrate your system at CNL 2012!

If you would like to participate, please register by sending an email to cnl2012@easychair.org not later than 10 August 2012. Your message should contain the following information: Your name, your affiliation and your email address.

There is no registration fee.

Practical information including a list of hotels and directions to the workshop venue can be found here:

http://attempto.ifi.uzh.ch/site/cnl2012/practical.html

We are looking forward to your participation at an exciting workshop.

Third Workshop on Controlled Natural Language (CNL 2012)

DEADLINE EXTENDED to 6 April

CONFIRMATION BY SPRINGER to publish proceedings within LNAI

29-31 August 2012
Zurich, Switzerland

http://attempto.ifi.uzh.ch/site/cnl2012/

A controlled natural language (CNL) is based on natural language but comes with restrictions on vocabulary, grammar, and/or semantics. The general goal is to reduce or eliminate ambiguity and complexity.

Some of these languages are designed to improve communication among humans, especially for non-native speakers of the respective natural language. In other cases, the restrictions on the language are supposed to make it easier for computers to analyze such texts in order to improve computer-aided, semi-automatic, or automatic translations into other languages. A third group of CNL has the goal to enable reliable automated reasoning on seemingly natural texts. Such languages have a direct mapping to some sort of formal logic and should improve the accessiblity of formal knowledge representations or specifications for people unfamiliar with formal notations.

All these types of CNL are covered by this workshop.

Topics

Possible topics for CNL 2012 include:

• CNL for knowledge representation
• CNL for question answering
• CNL for specifications
• CNL for business rules
• CNL for interactive systems
• CNL for machine translation
• CNL for improved understandability of texts
• design of CNLs
• CNL applications
• CNL evaluation
• usability and acceptance of CNL
• CNL grammars and lexica
• reasoning in CNL
• spoken CNL
• CNL in the context of the Semantic Web and Linked Open Data
• CNL in the government
• CNL in industry
• CNL use cases
• theoretical properties of CNL

Important Dates

Submission deadline (extended): 6 April 2012
Notification of acceptance: 25 May 2012
Deadline for revised papers: 18 June 2012
Workshop: 29-31 August 2012

Submissions and Proceedings

We invite researchers to submit papers with novel contributions in the area of CNL. These research papers should be formatted according to the Springer LNCS format and should not exceed 15 pages (but shorter papers are highly welcome too). Papers should be submitted in PDF format via the EasyChair conference system.

Accepted papers will be included in the printed workshop proceedings to be published by Springer within the LNAI series. Authors of accepted papers will be invited to present their research at the workshop. Unlike the previous CNL workshops, the final papers will be reviewed and published before the workshop (there are no extended abstracts anymore).

Venue

The workshop will take place at the Department of Informatics and the Institute of Computational Linguistics of the University of Zurich in Switzerland.

Organization Committee

• Tobias Kuhn (Yale University, USA), kuhntobias@gmail.com
• Norbert E. Fuchs (University of Zurich, Switzerland), fuchs@ifi.uzh.ch

Program Committee

• Johan Bos (University of Groningen, Netherlands)
• Peter E. Clark (Vulcan Inc, USA)
• Rogan Creswick (Galois, USA)
• Danica Damljanovic (University of Sheffield, UK)
• Brian Davis (DERI / National University of Ireland)
• Norbert E. Fuchs (University of Zurich, Switzerland)
• Normunds Gruzitis (University of Latvia)
• Stefan Hoefler (University of Zurich, Switzerland)
• Kaarel Kaljurand (University of Zurich, Switzerland)
• Peter Koepke (University of Bonn, Germany)
• Tobias Kuhn (Yale University, USA)
• Hans Leiss (University of Munich, Germany)
• Reinhard Muskens (Tilburg University, Netherlands)
• Gordon Pace (University of Malta)
• Richard Power (The Open University, UK)
• Laurette Pretorius (University of South Africa)
• Mike Rosner (University of Malta)
• Aarne Ranta (Chalmers University, Sweden)
• Rolf Schwitter (Macquarie University, Australia)
• Geoff Sutcliffe (University of Miami, USA)
• Silvie Spreeuwenberg (LibRT, Netherlands)
• Uta Schwertel (imc, Germany)
• Yorick Wilks (University of Sheffield, UK)
• Adam Wyner (University of Liverpool, UK)

Version 0.5.3 of AceWiki is available:

https://github.com/AceWiki/AceWiki/downloads

This release fixes a serious bug that has been introduced with version 0.5.2: the user interface crashed when a user wanted to edit a sentence.

Third Workshop on Controlled Natural Language (CNL 2012)

29-31 August 2012
Zurich, Switzerland

http://attempto.ifi.uzh.ch/site/cnl2012/

A controlled natural language (CNL) is based on natural language but comes with restrictions on vocabulary, grammar, and/or semantics. The general goal is to reduce or eliminate ambiguity and complexity.

Some of these languages are designed to improve communication among humans, especially for non-native speakers of the respective natural language. In other cases, the restrictions on the language are supposed to make it easier for computers to analyze such texts in order to improve computer-aided, semi-automatic, or automatic translations into other languages. A third group of CNL has the goal to enable reliable automated reasoning on seemingly natural texts. Such languages have a direct mapping to some sort of formal logic and should improve the accessiblity of formal knowledge representations or specifications for people unfamiliar with formal notations.

All these types of CNL are covered by this workshop.

Topics

Possible topics for CNL 2012 include:

• CNL for knowledge representation
• CNL for question answering
• CNL for specifications
• CNL for business rules
• CNL for interactive systems
• CNL for machine translation
• CNL for improved understandability of texts
• design of CNLs
• CNL applications
• CNL evaluation
• usability and acceptance of CNL
• CNL grammars and lexica
• reasoning in CNL
• spoken CNL
• CNL in the context of the Semantic Web and Linked Open Data
• CNL in the government
• CNL in industry
• CNL use cases
• theoretical properties of CNL

Important Dates

Submission deadline: 9 March 2012
Notification of acceptance: 25 May 2012
Deadline for revised papers: 22 June 2012
Workshop: 29-31 August 2012

Submissions and Proceedings

We invite researchers to submit papers with novel contributions in the area of CNL. These research papers should be formatted according to the Springer LNCS format and should not exceed 15 pages. Papers should be submitted in PDF format via the EasyChair conference system.

Accepted papers will be published in the printed workshop proceedings (probably by Springer). Authors of accepted papers will be invited to present their research at the workshop. Unlike the previous CNL workshops, the final papers will be reviewed and published before the workshop (there are no extended abstracts anymore).

Venue

The workshop will take place at the Department of Informatics and the Institute of Computational Linguistics of the University of Zurich in Switzerland.

Organization Committee

• Tobias Kuhn (Yale University, USA), kuhntobias@gmail.com
• Norbert E. Fuchs (University of Zurich, Switzerland), fuchs@ifi.uzh.ch

Program Committee

• Johan Bos (University of Groningen, Netherlands)
• Peter E. Clark (Vulcan Inc, USA)
• Rogan Creswick (Galois, USA)
• Danica Damljanovic (University of Sheffield, UK)
• Brian Davis (DERI / National University of Ireland)
• Norbert E. Fuchs (University of Zurich, Switzerland)
• Normunds Gruzitis (University of Latvia)
• Stefan Hoefler (University of Zurich, Switzerland)
• Kaarel Kaljurand (University of Zurich, Switzerland)
• Peter Koepke (University of Bonn, Germany)
• Tobias Kuhn (Yale University, USA)
• Hans Leiss (University of Munich, Germany)
• Reinhard Muskens (Tilburg University, Netherlands)
• Gordon Pace (University of Malta)
• Richard Power (The Open University, UK)
• Laurette Pretorius (University of South Africa)
• Mike Rosner (University of Malta)
• Aarne Ranta (Chalmers University, Sweden)
• Rolf Schwitter (Macquarie University, Australia)
• Geoff Sutcliffe (University of Miami, USA)
• Silvie Spreeuwenberg (LibRT, Netherlands)
• Uta Schwertel (imc, Germany)
• Yorick Wilks (University of Sheffield, UK)
• Adam Wyner (University of Liverpool, UK)

AceWiki version 0.5.2 is available:

https://github.com/AceWiki/AceWiki/downloads

The AceWiki repository moved from Launchpad to GitHub.

The most important change is that the core interfaces of the internal structure have been redesigned in order to support multilinguality. There is now an experimental GF engine as an alternative to the ACE-OWL engine: It uses GF (Grammatical Framework) to allow for automatic translations into other languages (the current prototype uses a very simple grammar and is still very buggy). Furthermore, AceWiki applications can now be split into backends and frontends: Multiple frontends can connect to the same backend, i.e. to the same ontology. A change with practical importance is that jpl.jar (Java binding for SWI Prolog) is no longer part of the AceWiki package but has to be linked dynamically. Apart from that, there are many minor bug-fixes and improvements, as well as general refactoring and clean-up. See the change log for the details.

The source code of APE (the ACE parser) is now hosted on Github:

https://github.com/Attempto/APE

A new release of the Attempto Parsing Engine (APE) is available for download:

http://attempto.ifi.uzh.ch/site/downloads/

This is a bugfix release that brings minor changes and updates:

• ACE to DRS translation:
  • fixed: sentences like "For every woman of who everybody waits." now correctly fail
  • fixed: a bug that in rare cases generated a DRS with duplicate referents
  • fixed: duplicate DRS conditions are now always suppressed
  • fixed: misguiding error message for imperatives
  • deprecated: the construct "the ... of which/who"
• DRS to FOL translation:
  • fixed: disjunctions related by anaphoric references (e.g. the DRS of "A man sleeps or the man waits.") are now correctly mapped to FOL
• DRS to ACE translation:
  • improved: handling of disjuntions, e.g. "Every dog barks or a cat sleeps." is now paraphrased as "A cat sleeps or if there is a dog X1 then the dog X1 barks."
  • restored: support for 0, e.g. "There are 0 men."
• DRS to OWL/SWRL translation:
  • the pretty-printed output (owlfsspp) now conforms to the OWL Functional-Style Syntax (and can be parsed by the OWL-API)
  • avoid including redundant owl:Thing in intersections
  • now supported: the DRS of "John does not see what?"
  • fixed: sentences like "There is X. If X is a fireman then X is a man." now map to OWL (instead of SWRL as before)
  • improved: the SWRL-mapping now supports formula/3 with =<, >=, \=
  • fixed: "Peter owns 3 kg of apples." now correctly fails
• Attempto Java Packages:
  • updated the Apache HTTP libraries to httpcore-4.1.1.jar and httpclient-4.1.1.jar
  • removed the OWL Verbalizer Webservice interface, this is now available at owlverbalizer.googlecode.com
  • updated enum OutputType: removed OWLRDF, added TPTP
• The APE regression test set (~3500 small ACE texts with their corresponding DRSs) is now included (see ape/acetext_drs.pl)
• Several improvements to the documentation of ACE, DRS and the APE source code.

The Attempto Group, working for you in Barcelona at the GF Summer School.

Release of AceWiki version 0.5.1 is available:

https://launchpad.net/acewiki

Besides internal changes in the package structure, the most important change is that users have the option to stay logged in, elimitating the need to provide the password again the next time if AceWiki is accessed from the same browser instance. Apart from that, there are many small changes, new configuration parameters, GUI adjustments, and bug fixes. See the changelog for the details.

AceWiki version 0.5.0 has been released:

https://launchpad.net/acewiki

It uses version 3 of the Echo Web Framework, instead of version 2. This major upgrade should make AceWiki more stable and ready for the future development. Other third-party libraries have been updated too. Another major change concerns the predictive editor: The menu blocks are arranged in a smarter way and have different colors to emphasize their types. As usual, a number of bugs have been fixed and there has been clean-up and refactoring on the code. See the changelog for the details.

Version 0.4.4 of AceWiki has been released:

https://launchpad.net/acewiki

This is mainly a bug-fix release. Some bugs sneaked in at the last release, others have been around for longer. See the changelog for the details.

Version 0.4.3 of AceWiki has been released:

https://launchpad.net/acewiki

This release includes several changes in the user interface of AceWiki. The word information is accessed by clicking on the article title instead of word tabs. The main page is now a wiki article too. Furthermore, the search field has moved to the top right. Additionally, there are many internal improvments. See the changelog for the details.

A new release of the Attempto Parsing Engine (APE) is available for download (version 6.6-101116):

http://attempto.ifi.uzh.ch/site/downloads/

The changes are:

• The tokenizer has been improved. It is now faster and some minor bugs have been fixed.
• The Prolog operators [] and <> used to express necessity and possibility have been replaced by 'must' and 'can'. This is due to the fact that recent versions of SWI Prolog prohibit the use of [] as an operator.
• "such that" is deprecated and generates a warning message. The reason is that the interpretation of "such that" in ACE does not match its natural use in English.
• OWL/SWRL output has been improved:
  • OWL 2 instead of OWL 1.1.
  • Better SWRL support.
  • Some OWL axioms are simplified.
  • RDF/XML output is now deprecated, one can use alternative solutions to produce it, e.g. OWL-API, Thea.
  • Support for positive attributive and predicative adjectives: they translate into OWL classes.
• The Attempto Java Packages are now part of the APE package.

ACE View v1.3.0 is now available from

http://attempto.ifi.uzh.ch/aceview/

Compared to ACE View 1.2.x there are few new features, most of the changes are there to make ACE View compatible with Protege 4.1. See the release notes at:

http://aceview.googlecode.com/svn/trunk/RELEASE_NOTES.html

Version 0.4.2 of AceWiki has been released:

https://launchpad.net/acewiki

The most important change of this release is the support for user login. The login is password protected and users can register themselves. Apart from that, the graphical user interface has been improved. For example, the buttons "Back", "Forward" and "Refresh" now have icons. Furthermore, there are several minor changes. See the changelog for the details.

AceWiki version 0.4.1 is available for download:

https://launchpad.net/acewiki

This release solves several problems. Most importantly, two memory leaks have been fixed (memory leaks in the broad sense of the term). References to closed windows were internally retained, which prevented them from being garbage collected. The second memory leak was due to the fact that references to the closed application instances were retained. Additionally, there is a new AceWiki data format, in which the wiki content can be exported (on the client-side) and imported (on the server-side). Furthermore, this is the first release that includes the test scripts for the two grammars.

A new version of AceWiki (0.4.0) is available on Launchpad:

https://launchpad.net/acewiki

See the AceWiki page for more information and the public demo instances:

http://attempto.ifi.uzh.ch/acewiki/

This release contains major updates on the reasoner integration. Most importantly, the reasoner is now exchangeable and external reasoners can be connected. Version 3 of the OWL API is now used (instead of version 2) and new parameters are available when launching an AceWiki instance. Furthermore, there is a new export option to download the wiki content as two CSV lists. And there are many more changes; see the change log for more details.

A new release of the Attempto Parsing Engine (APE) is available for download (version 6.5-100128):

http://attempto.ifi.uzh.ch/site/downloads/

There are only few changes:

• For consistency and convenience reasons, zero is now supported in cases like "John has 0 cars." and "John has more than 0 cars.". However, there are better alternatives than using "0" in most cases (e.g. "John has no cars.", "John has at least 1 car.").
• Strings can now contain quotation marks, which have to be escaped by a backslash, e.g. "John types "ab\"cd".". Backslashes have to be escaped too: "John types "ab\\cd".".
• Some complex but incorrect sentences took very long or even encountered a timeout. This is now fixed.
• In some rare cases, function words were interpreted as prepositions for prefixed transitive adjectives or ditransitive verbs. This is also fixed.

The source code of AceWiki and the ACE Editor is now hosted on Launchpad:

• AceWiki (and the ACE Editor) on Launchpad

Bugs and questions regarding AceWiki or the ACE Editor can now be posted there. Furthermore, code contribution is highly welcome.

We are happy to announce a new release of the Attempto Java Packages, which include AceWiki and the ACE Editor.

The most important changes concern the grammar notation (Codeco) and the chart parser. See [1] for details. Also the ACE Editor and AceWiki have been improved.

• General
• The code has been cleaned up, refactored, and documented
• Many internal classes have been changed in a non-backwards-compatible way
• The latest version of Echo2 is used
• Codeco
• The Codeco notation has been extended and improved (see [1])
• Codeco-to-LaTeX transformation
• Codeco-to-HTML transformation
• Wellformedness checker for Codeco
• Chart Parser
• Improved parsing algorithm: it is now faster and several bugs have been removed
• Redesigned lookahead features
• Predictive Editor
• The resolution of anaphoric references is visualized by highlighting antecedent and anaphor
• ACE Editor
• Improved grammar (based on Codeco)
• New words can be added within the predictive editor (as in AceWiki)
• Lexicon files can be saved and loaded
• The start-up parameter "immutable_lexicon" can be used to disallow lexicon changes
• Nicer buttons in the main window
• AceWiki
• Improved grammar (based on Codeco)
• The triangle symbols have now tool tips
• Less flashy icons are used to indicate the recalculation of reasoning results
• Better synchronization of modification and query tasks
• Some bugs have been fixed

The Attempto Java Packages can be downloaded from our download page:

http://attempto.ifi.uzh.ch/site/downloads/

[1] Tobias Kuhn. Controlled English for Knowledge Representation. Doctoral thesis FINAL DRAFT, Faculty of Economics, Business Administration and Information Technology of the University of Zurich, to appear. http://attempto.ifi.uzh.ch/site/pubs/papers/doctoral_thesis_kuhn.pdf

A new release of the Attempto Parsing Engine (APE) is available for download (version 6.5-091130):

http://attempto.ifi.uzh.ch/site/downloads/

Though we have many ideas and also requests to extend Attempto Controlled English, this release serves mainly to consolidate the current status and contains no new language features.

The Attempto Parsing Engine (APE) was tested extensively which revealed a few bugs that we subsequently fixed:

• Adjectives used with the construct "as ... as" were sometimes represented incorrectly.
• Rare forms of combining relative clauses and anaphoric references did not work.
• Scopes were represented incorrectly in some rare cases.
• For some questions, parsing failed but succeeded if a full stop instead of a question mark was used.
• In some rare cases, parsing succeeded for irreflexive pronouns where reflexive pronouns should have been used.
• In some cases, proper names with relative clauses did not work properly with anaphoric references.
• There were several small bugs in the resolution of anaphoric references.
• Unknown words were not reported if the first sentence succeeded.

Also we cleaned up the code of APE and improved its internal documentation.

APE's paraphraser was slightly amended:

• It now correctly chooses "a" or "an".
• If the paraphraser cannot generate a paraphrase ‒ for instance for distributed plurals, some questions, some cases of sentence subordination ‒ it generates the ACE comment "# ..." instead.

APE's TPTP output:

• The TPTP output was improved.
• Several bugs were fixed.

In brief, APE is now in a very stable state.

A new release of the Attempto Parsing Engine (APE) is available for download (version 6.5-090616):

http://attempto.ifi.uzh.ch/site/downloads/

The changes are the following:

• Two bugs concerning the XML representation of the DRS have been fixed:
  • The contents of the XML attributes were not properly escaped.
  • For some sentences, the creation of the XML representation failed altogether.
• There is a new internal module that splits a DRS into a list of smaller DRSs. This makes the creation of several output types more efficient and more robust:
  • For the "paraphrase" output, APE splits now the DRS, verbalizes each of the resulting smaller DRSs in NP ACE and if that fails then in Core ACE, and finally merges the verbalizations and presents them as separate paragraphs. Note that the output is a mixture of Core ACE and NP ACE. Paragraphs that could not be paraphrased appear as a comment containing three dots "...".
  • For the "paraphrase1" output, APE also uses DRS splitting, but the output is always Core ACE. The "paraphrase2" output is unchanged.
  • The ACE to OWL translation uses now DRS splitting, translates each of the resulting smaller DRSs into OWL, and finally merges these OWL axioms. This has the consequence that there are fewer anonymous individuals in some cases.

Only hours after the release yesterday, we found a medium-sized bug in the Attempto Parsing Engine: Warning messages were returned only for the last paragraph. This is now fixed and a new package of APE is available for download. Sorry for the inconvenience.

We are releasing version 6.5 of Attempto Controlled English (ACE) and new versions of the Attempto Parsing Engine (APE) and the Attempto Java Packages.

Changes of Attempto Controlled English (ACE)

• ACE provides now modality with modal auxiliaries for recommendation ("should"), and admissibility ("may"). These forms of modality can also be expressed by the fixed phrases "it is recommended that", respectively "it is admissible that".
• Declarative sentences, questions and commands can now be arbitrarily mixed. All sentences can anaphorically refer to preceding declarative sentences, but not to questions or commands.
• If in sentence subordination the main sentence and the subordinated sentence have the same subject as in "John wants that he sleeps" then this sentence can be more concisely expressed as "John wants to sleep".
• We introduce the notion of paragraphs for ACE texts. An empty line (or a line that contains only white spaces) is interpreted as a paragraph break. Anaphoric references are only possible within the same paragraph.
• All verb phrases including the copula can now consistently be modified by adverbs and by prepositional phrases. These modifiers can now also precede the verb phrase, for example "John directly flies to Zurich.".
• In order to be more flexible with respect to the many ways plurals can be interpreted, all noun phrases with generalised quantifiers are now accessible for anaphoric references. For example: "John has at most 2 cars. The cars are expensive."
• Commands can now contain negation, for example "John, do not touch the machine!".
• Macros like for example "Proposition X: John waits. It is true that X." have been removed from the language.
• Negated generalized quantifiers like for example "not more than 5" have been removed from the language because they are harder to understand than equivalent generalized quantifiers without negation (the given example is equivalent to "at most 5").
• Sentences like "there is X" are now supported and are interpreted as "there is something X"
• Commands can now use the correct second person pronouns "you", "your", "yourself", "yourselves", and "your own".
• Relative clauses can be used only for nouns, indefinite pronouns and proper names.

Changes of the Attempto Parsing Engine (APE)

• DRS:
  • Proper names no longer introduce separate conditions in the DRS. Instead, a term of the form "named(...)" is introduced for each occurrence of the proper name. For example, "John likes Mary" introduces now only one DRS condition: predicate(A, like, named('John'), named('Mary')).
  • The modal constructs for recommendation ("should", "it is recommended that") introduce nested DRS boxes using the operator "should". In the same way, admissibility ("may", "it is admissible that") introduces the operator "may".
  • Questions introduce now nested DRS boxes using the operator "question" and in the same way commands use the operator "command".
  • The DRS conditions have now not only a sentence id attached but also a token id that points to the position of the token within the sentence.
  • Maximality conditions (e.g. for "at most 5") no longer introduce implications in the DRS but use plain Prolog lists to bracket the conditions that are within the scope of the maximality.
• Paraphraser:
  • Some additional forms of that-subordination and questions work now.
• OWL:
  • Anonymous individuals have now the prefix "Ind" both in XML and in RDF (used to be "I" and "id").
  • Less anonymous individuals are created. For example, no anonymous individuals are created for "John is a man that is a manager".
  • The auxiliary entities "Superman" and "know" are replaced by "Universe" and "contain".
• TPTP:
  • APE can now also translate ACE texts into the TPTP format (see http://www.tptp.org).

Changes of the ACE Lexicon Format

• Nouns and proper names can now have undefined gender. In this case, non-human (e.g. "it") or human (e.g. "her") pronouns can be used.

Changes of the Attempto Java Packages

• The Attempto Java Packages contain now the code for the ACE Editor.
• The ACGN grammar format (that is used for the chart parser and the predictive editor) has been replaced by the new format Codeco (COncrete and DEclarative grammar notation for COntrolled natural languages). Anaphoric references can now be defined in a fully declarative way. (This change is not backwards compatible if you defined your own grammar and/or menu creator with an earlier version of the Attempto Java Packages.)
• The class HeaderControlFilter has been removed. It was a work-around for a bug in Firefox 3.0.0 to 3.0.4 and now we can assume that nobody uses these versions of Firefox anymore.
• AceWiki has now client-side support for the export of the ACE lexicon and of the whole knowledge base as an ACE text.
• Comments in AceWiki can now contain wiki-internal links (using double square brackets, e.g. "[[country]]") and external links using URLs.

Furthermore, many minor improvements and bug-fixes have been done. The latest versions of APE and the Attempto Java Packages can be downloaded from our download page:

http://attempto.ifi.uzh.ch/site/downloads/

Feedback is always appreciated.

We are happy to announce the public demo of the ACE Editor:

http://attempto.ifi.uzh.ch/aceeditor/

The ACE Editor demonstrates how editing of ACE texts can be done in a convenient way. It uses the predictive editor that is also used in AceWiki (but the ACE Editor supports a larger subset of ACE than AceWiki does). The ACE Editor is not a finished tool but rather a general basis to create domain-specific tools on top of it. The code will be made public as a part of the Attempto Java Packages in the near future.

The address for the Attempto mailing list has changed. The new address is:

attempto@lists.ifi.uzh.ch

From now on, this new address should be used. The old one (attempto@ifi.uzh.ch) will still work for the next couple of months.

The Attempto project has now a mailing list:

attempto@ifi.uzh.ch

The subscription form and further information can be found on the Attempto mailing list page. This list is intended for questions, comments, bug reports, announcements, etc. concerning the Attempto project and its languages and tools. Everyone can subscribe to the list and everyone can send messages to it (we manually filter out spam).

Today, the semantic wiki engine AceWiki has been released as open source and is now a part of the Attempto Java Packages. See our download page and the AceWiki website. AceWiki has left its prototype status and is now available as alpha version 0.2.8.

Website: http://attempto.ifi.uzh.ch/site/cnl2009/
Submission deadline: 14 November 2008
Workshop date: 8-10 June 2009
Location: Marettimo Island, Italy

Controlled Natural Languages (from Wikipedia):

Controlled natural languages (CNLs) are subsets of natural languages, obtained by restricting the grammar and vocabulary in order to reduce or eliminate ambiguity and complexity. Traditionally, controlled languages fall into two major types: those that improve readability for human readers, and those that enable reliable automatic semantic analysis of the language. [...] The second type of languages has a formal logical basis, i.e. they have a formal syntax and semantics, and can be mapped to an existing formal language, such as first-order logic. Thus, those languages can be used as knowledge representation languages, and writing of those languages is supported by fully automatic consistency and redundancy checks, query answering, etc.

Various controlled natural languages of the second type have been developed by a number of organisations, and have been used in many different application domains, most recently within the semantic web.

This workshop is dedicated to discussing the similarities and differences of existing controlled natural languages of the second type, possible improvements to these languages, relations to other knowledge representation languages, tool support, existing and future applications, and further topics of interest.

The plain text version of the call for submissions can be found here.

Topics

CNL 2009 will address the following aspects of controlled natural languages (CNLs):

• design of CNLs
• parsing of CNLs
• CNLs for knowledge representation
• CNLs for specifications
• CNLs and the semantic web
• CNLs as user interface
• CNLs for interaction and communication
• tool support for CNLs
• reasoning in CNLs
• comparisons of CNLs
• applications of CNLs
• business rules
• user studies
• theoretical results
• etc.

The workshop will be informal with lots of time for presentations and discussions in the fashion of the seminars organised at Dagstuhl in Germany. To ensure the informal atmosphere the number of participants will be limited.

Submission Details

We invite researchers to submit extended abstracts of exactly 4 pages (inclusive references). These extended abstracts will be intensively reviewed by several members of the programme committee. Authors of accepted extended abstracts will be invited to present their research at the workshop. Revised versions of the accepted abstracts will be published before the workshop as a technical report of the Department of Informatics of the University of Zurich. During the workshop authors will have one full hour to present their work and to have it discussed by the participants. All authors are then invited to submit a full paper of up to 20 pages (inclusive references) that takes the discussions during the workshop into account. Full papers will again be reviewed by the programme committee. Revised versions of the full papers will be published by Springer in their LNCS/LNAI series.

Extended abstracts and full papers should use the Springer LNCS format. Extended abstracts must be submitted electronically in PDF format. For paper submissions we use EasyChair.

Venue

The workshop will take place on the Italian island Marettimo at the Marettimo Residence that consists of a set of two-storey houses within a beautiful garden. On top of the garden of the residence there is a large lecture hall with wireless internet.

Marettimo is the outermost of the Egadian Islands to the west of Sicily, and is easily reached from the airports of Palermo and Trapani. Marettimo offers the simple and relaxed life of southern Italy, unspoilt landscape, stupendous views, hiking, swimming, diving, boat trips, and excursions on donkeys. There are several restaurants and bars, and some shops. What the island does not offer: traffic - there are practically no roads - fancy shops and restaurants, night life, and sandy beaches.

Accommodation

The Marettimo Residence offers one- and two-bedroom apartments with fully equipped kitchens. A number of apartments for the participants of CNL 2009 will be reserved until early 2009 at a price 10% below the regular price. Participants should get into direct contact with the Marettimo Residence to organise their accommodation. Alternative accommodations on Marettimo can be found via the internet.

Conference Dinner

A conference dinner will be arranged during the workshop, and will be paid individually by the participants.

Important Dates

• Deadline for submissions of extended abstracts: 14 November 2008
• Notification of acceptance of extended abstracts: 9 January 2009
• Participants contact Marettimo Residence concerning accommodation: end of January 2009
• Revised versions of extended abstracts: 13 March 2009
• Workshop: 8-10 June 2009
• Deadline of submission of full papers: 4 August 2009
• Feedback on full papers: 2 October 2009
• Revised versions of full papers: 6 November 2009

Program Committee

• Piero Bonatti (University of Naples, Italy)
• Johan Bos (University of Rome "La Sapienza", Italy)
• Peter E. Clark (Boeing, Seattle, USA)
• Hamish Cunningham (University of Sheffield, UK)
• Enrico Franconi (University of Bolzano, Italy)
• Norbert E. Fuchs (University of Zurich, Switzerland) (chair)
• Glen Hart (Ordnance Survey, Southampton, UK)
• Jerry R. Hobbs (USC/ISI, USA)
• Kaarel Kaljurand (University of Zurich, Switzerland)
• Peter Koepke (University of Bonn, Germany)
• Tobias Kuhn (University of Zurich, Switzerland)
• Ian Pratt-Hartmann (University of Manchester, UK)
• Stephen Pulman (University of Oxford, UK)
• Mike Rosner (University of Malta, Malta)
• Rolf Schwitter (Macquarie University, Australia)
• John Sowa (VivoMind, USA)
• Silvie Spreeuwenberg (LibRT, Amsterdam, Netherlands)
• Uta Schwertel (imc, Germany)
• Yorick Wilks (University of Sheffield, UK)

Organisation

Norbert E. Fuchs (University of Zurich, Switzerland), fuchs@ifi.uzh.ch

We are releasing the Attempto Java Packages and a new version of APE under the GNU Lesser General Public License:

http://attempto.ifi.uzh.ch/site/downloads/

The Attempto Java Packages is a collection of Java packages that are related to the Attempto project. There are four jar-files:

• attempto_ape.jar contains different interfaces to the ACE parser that make it easy to use APE from Java programs.
• attempto_chartparser.jar contains a chart parser (concretely an Earley parser) that is used by AceWiki.
• attempto_echo.jar contains different graphical components using the Echo Web Framework, most prominently a predictive editor that is used in AceWiki.
• attempto_owl.jar contains an interface to the OWL verbalizer webservice

Documentation of the Attempto Java Packages can be found here:

http://attempto.ifi.uzh.ch/site/docs/java/

There are some small changes in the language ACE that are a first step towards the next version 6.1 of ACE where these changes will be properly documented:

• Support for bare variables: Variables can now be introduced by themselves, e.g. besides "a person X" or "somebody X" also just "X" can be used to introduce a new variable.
• Restricted variable names: The names of variables have to follow the pattern "[A-Z][0-9]*", i.e. an uppercase character followed by zero or more digits.
• Support for proper names with definite articles: There is now a new class of proper names requiring a definite article, e.g. "the United-Nations", "the Nile", "the University-of-Zurich", etc.

Apart from those extensions of the language, the parser has been improved in several ways:

• The resolution of definite noun anaphors is more liberal. In addition to resolving the anaphor "the man who owns a dog" from the antecedent "a man who owns a dog", the anaphor can now also be resolved from the more natural antecedent "a man owns a dog".
• Error messages point now to the position within the sentence where parsing failed.
• The executable "ape.exe" can now also act as a server.
• Some bugs have been fixed.
• Performance and stability improvements have been achieved.

Finally, a new lexicon format has been defined. It is now the same format that is used internally for the source file "clex_lexicon.pl" (the larger version of this lexicon is also updated and available for download). The older formats are still supported. The most obvious change is that prepositions are not built-in function words anymore but can be defined in the lexicon. This is the specification of the new format:

http://attempto.ifi.uzh.ch/site/docs/ace_lexicon.html

Any kind of feedback is appreciated.

Today we are announcing a new version of the Attempto Reasoner RACE. This version covers most of ACE 6, but does currently not support modality, sentence subordination and operations on arithmetic expressions, strings, lists and sets.

RACE offers three modes of deductions:

• consistency checking of ACE axioms
• proving ACE theorems from ACE axioms
• answering ACE questions from ACE axioms

For each mode of deduction, RACE finds all minimal subsets of the axioms involved, and displays them in ACE.

RACE answers also the question "why not?": if a proof or query answering fails then RACE will display those parts of the theorems/query that could not be deduced.

Deductions are supported by auxiliary axioms that encode domain independent knowledge, for instance the relations between plural and singular nouns, or the relations between natural numbers.

We applied several means to speed-up RACE. As a result, the new version is noticeably faster than its predecessor.

RACE can be accessed via its web service

http://attempto.ifi.uzh.ch/site/docs/race_webservice.html

or via its web client

http://attempto.ifi.uzh.ch/race/

that provides a graphical front-end to the web service.

We appreciate your feedback.

Today we are releasing the source files of the Attempto Parsing Engine (APE) plus some related tools under the GNU Lesser General Public License. Download the zip-archive from:

http://attempto.ifi.uzh.ch/site/downloads/files/ape-6.0-080506.zip (228 KB)

(Note that all the downloads are also available from the new Downloads-section on our website.)

These tools are completely written in Prolog, specifically SWI-Prolog (http://www.swi-prolog.org). There are no other dependencies. In order to use APE, you must install SWI-Prolog and its packages (minimally: clib and sgml) first. Recent versions of SWI-Prolog (5.6.46 and higher) should be OK. SWI-Prolog is available in a pre-compiled form for Windows, Mac OS X and many Linux distributions. Its source code is also available. See http://www.swi-prolog.org/dl-stable.html. Please let us know if you encounter problems using APE on your platform.

The distribution includes the following packages:

• parser/ contains the Attempto Parsing Engine (APE) (tokenizer, grammar files, anaphoric reference resolver)
• lexicon/ contains various lexicon files, notably a content words lexicon with ~2,000 entries
• utils/ contains various modules, mostly for translating the Discourse Representation Structure (DRS) generated by APE into other forms (standard first-order logic syntax, OWL/SWRL, back to ACE, etc.)
• logger/ contains the error logger module

There is an easy-to-use command-line tool `ape.exe' that provides an interface to APE and the other modules. For example, executing:

./ape.exe -text "Every man is a human." -solo drsxml

will output the DRS of the ACE sentence "Every man is a human." in XML.

To compile this command-line tool, just execute `make_exe.sh' (or `make_exe.bat', if you are using Windows).

In a separate package we are also releasing a content words lexicon with ~100,000 entries under the GNU General Public License. This lexicon can replace the smaller content word lexicon that comes with the APE package. Download the zip-archive from:

http://attempto.ifi.uzh.ch/site/downloads/files/clex-6.0-080506.zip (668 KB)

and just replace `lexicon/clex.pl' by `clex.pl' contained in this archive.

There is more to come. The next release will include more support for the developers: more documentation of the various modules, our regression test set with some 3000 ACE-DRS pairs, example code, instructions on how to set up a web-interface for APE, how to call APE from Java, etc.

In the future we plan to make even more ACE tools public as open source. These tools include the ACE reasoner RACE, the AceRules system, the AceWiki, the ACE View ontology editor, the OWL verbalizer.

Please contact us for support, feature requests, alternative licensing options, etc. by sending an email to fuchs@ifi.uzh.ch.

Today we are announcing Attempto Controlled English 6.0.

Compared to the previous version 5.5 there are major extensions and a few minor restrictions to the syntax of ACE.

Extensions to the syntax:

• ACE 6 supports compound expressions, lists and sets, and formulas. Some examples:
  • A value is (1 + 2) / X * 4.
  • "abc" & "123" is a valid password.
  • 3 is the first element of [3, 4, 5, "ab", John, 1+2].
  • {3, 6, [1,2]} contains 6.
  • 10 = 4 + 6.
  • 5 > 3.
  • X >= 13.4.
  • 3 < 4 and 3 =< 5.
  Note that expressions and formulas are not evaluated by the parser.
• Support was added for the determiner phrase `nothing but' (`nobody but') and its generalisation `no N but', where N is a common noun (e.g. `no man but'). This determiner phrase can be used with plural and mass noun phrases and with proper names.
  • Every koala eats nothing but eucalyptus-leaves.
  • Mary likes nobody but John.
  • Mary likes no man but John.
  • John likes no cars but trucks.
  • No man but John waits.
• A new generalised quantifier `exactly' was introduced.
  • Every dog has exactly 4 legs.
  • John has not exactly 1 car. (= John does not have exactly one car.)
• The syntax for questions was extended, e.g. the following questions are now allowed:
  • Who does John not see?
  • What does a man not eat?
  • Where does a man not eat?
  • There is who?
  • There is which man?
  • Where is John?
  • Whose dog barks?
  • Who waits?
  • Who wait? (i.e. the query word can be used in plural context)
  • Each of who waits?
  • John is a father of who?
  • Bill gives what to who?
  • Is it true that a man waits?
  • Is it false that a man waits?
• Saxon genitive is now also allowed with indefinite pronouns and variables, e.g.
  • There is a man X. X's dog barks.
  • Everybody's dog barks.
  • For every man's cat a dog barks.
• Integers and reals can be followed by a measurement noun. The symbols of SI base units (`m', `kg', `s', `A', `K', `mol', `cd', ...) are included as predefined measurement nouns in APE's built-in lexicon.
  • D is a distance from Earth's equator to North-Pole. 1m is D/10000000.
  • John's height is 2 m.
• Measurement constructs can be both singular and plural:
  • 2 kg of water flows. (= 2 kg of water flow.)
• The `each of' construct can now be used in relative clauses:
  • There are some men each of who waits.
• Support was added to `all'/`no' + plural noun phrase. This is interpreted as `every'/`no' + singular noun phrase, e.g.
  • All cats sleep. (= Every cat sleeps.)
  • No dogs sleep. (= No dog sleeps.)

Restrictions to the syntax:

• The support for variable conjunction in apposition was dropped, e.g. "John sees 2 men X and Y." is now parsed so that `Y' is taken to be a proper name.
• The support for `it is not the case that' was dropped, use `it is false that' instead.
• Adverbs are not allowed to modify copula verbs anymore.
• Noun phrase conjunctions used in `there are' constructs cannot contain definite NPs, proper names, and numbers/strings anymore.
• Noun phrases introduced by generalised quantifiers `at most' and `less then' are not accessible to anaphoric references anymore.

As ACE 6.0 is more expressive than ACE 5.5, several changes were made to the DRS representation of ACE sentences.

• Representation of expressions and formulas was introduced.
• Representation of `at most' and `less than' was changed.
• Representation of numbers and strings was simplified.

Support for ACE 6.0 has been implemented in APE (the Attempto Parsing Engine). Additionally:

• Error reporting was extended and improved.
• There are many performance improvements and some bugfixes.
• The Syntax Report that presents the formal grammar used by APE has been cleaned up.
• Support for Unicode was fixed. (In ACE, Unicode symbols encoded in UTF8 can be used in quoted strings.)
• Saxon genitive can now be used with unknown proper names, e.g. "Jimmy's dog barks.".

Verbalisation of DRSs:

• Paraphrasing of some forms of query-conditions was introduced.

For more details, read the updated documentation.

The new features are immediately available as part of the Attempto web-interface and the APE web-service. The previous version of the APE web-service stays temporarily accessible at: http://attempto.ifi.uzh.ch/ws/ape-old/apews.perl

We appreciate your feedback!

Today, version 2 of AceWiki has been released. The main improvements are:

• Now the complete ontology is expressed in ACE.
• Better usability
• Improved authoring tool
• Wizards for creating new entities
• Menus in the wiki pages
• ACE parser integration
• OWL/SWRL support

The URL of the new AceWiki is:

http://attempto.ifi.uzh.ch/acewiki/

Documentation will follow. As always, feedback is very welcome.

Today the AceRules webservice has been updated. It is now fully compliant with SOAP 1.1 and WSDL. Furthermore, importing a user lexicon is now supported.

For the details, check out the AceRules webservice specification.

The two web interfaces for AceRules have been updated accordingly:

• AceRules Interface
• AceRules Technical Interface

The new webservice is not compatible with the old one. Thus, programs using the webservice need to be changed. The old webservice will stay accessible for a while at:

http://attempto.ifi.uzh.ch/ws/acerules-old/acerulesws.perl

Today we are announcing Attempto Controlled English 5.5.

New language features:

• numbers and strings as general objects, e.g. "John lives in "Paris".", "The temperature reaches -2.", "3.14 badly approximates Pi."
• positive, comparative and superlative adjectives, e.g. "John is tall.", "John is as tall as Mary.", "Mary is taller.", "Mary is taller than John.", "Mary is tallest."
• positive, comparative and superlative adverbs, e.g. "John runs fast.", "Mary runs faster.", "Mary runs fastest."
• simple form of imperatives, e.g. "John, enter a card!" (translated as "John enters a card.")
• generalisation: any noun phrase can now take a relative phrase
• restriction: no can only precede singular noun phrases, no longer plural noun phrases

Lexicon:

• clean-up of the built-in lexicon
• new lexicon format
• improved error handling for the user lexicon

Attempto Parsing Engine:

• extended error reporting
• improved error representation in the web-interface

Verbalisation of DRSs:

• DRS verbalisation component now uses the ACE lexicon
• improved paraphrase of ACE input

DRS Representation:

• major overhaul leading to simplifications
• typed and untyped representations
• extended representation of adjectives and adverbs

ACE-to-OWL Translation:

• added support for OWL 1.1 (data properties partially supported)
• two outputs provided: OWL 1.1 functional-style syntax and OWL 1.1 RDF/XML syntax

For more details, read the updated documentation.

The new features are immediately available as part of the Attempto web-interface and the APE web-service. The previous version of the APE web-service stays temporarily accessible at: http://attempto.ifi.uzh.ch/ws/ape-old/apews.perl

As always, we appreciate your feedback.

We have updated the APE webservice today.

1. The new webservice includes performance enhancements: in case of short input texts the service now responds up to 3x faster.
2. APE also offers more and better error messages, e.g.
   • missing determiner, e.g. in "John owns car."
   • undefined propername, e.g. "Peeter owns a car."
   • malformed file, in case the user lexicon doesn't conform to the specification
3. APE does no longer generate a warning message when possessive pronouns (his, his own, whose, ...) are used.
4. The APE webclient now offers only one type of paraphrase while the webservice continues to offer both previously available types of paraphrase.

Furthermore, the documentation section of the Attempto website has three new documents:

• Authoring Tools for ACE discusses the use of predictive authoring tools for ACE
• OWL ACE Construction Rules lists the rules that determine which sentences belong to the subset of ACE that is used when verbalizing OWL ontologies
• ape_lexicon.txt is a list of words from APE's built-in lexicon, classified into word classes

Let us know in case you encounter any problems.

There will soon be another release of ACE, offering some enhancements of the language ACE itself, an additional simplified DRS representation, and more. Stay tuned!

The biggest news is that we have a dedicated server which offers much faster access to the ACE tools. In particular, APE is about 5 times faster now.

We are also announcing a set of new ACE-based tools.

AceRules is a forward-chaining rule system based on ACE, i.e. the rules and the answer sets are written in ACE. It supports multiple semantics (so called "modes"): courteous logic programs, stable models, and stable models with strong negation. There is a public webservice for AceRules. See the document AceRules Webservice for more information. Based on this webservice, there are two web interfaces: the AceRules interface as a demonstration for normal users and the AceRules technical interface for developers.

AceWiki is a semantic wiki using ACE. It has a preloaded ontology of protein knowledge. Feel free to extend and change the wiki content. Please note that AceWiki is still at an experimental state: the wiki content is not stored persistently, and there is not much documentation. See AceWiki for more.

OWL verbalizer converts OWL DL (expressed in RDF/XML) into ACE. At the moment, data-valued properties are not supported. See the document Verbalizing OWL in (controlled) English for more.

Other changes

• A new directory Courses contains the slides and exercises of an ACE course that we gave at the University of Zurich in December 2006.
• A draft of Syntax Report (the grammar used by APE) is now available.
• The APE webservice now optionally returns the DRS in XML (specification forthcoming).
• The set of intransitive adjectives in APE's built-in lexicon was supplemented with 5734 past participles (e.g. validated, x-rayed).
• There are some fixes and extensions to the ACE paraphraser.

As we have moved to a new server, the URLs of our website and tools have changed. The old URLs will be deactivated after a month. Please update your bookmarks. The new URLs are:

• Attempto website: http://attempto.ifi.unizh.ch/site
• APE webclient: http://attempto.ifi.unizh.ch/ape
• APE webservice: http://attempto.ifi.unizh.ch/ws/ape/apews.perl

Today we are announcing ACE 5. We have extended ACE 4 by:

• modality, e.g. "A customer must enter a code.", "It is possible that every service crashes."
• sentence subordination (that-subordination), e.g. "John believes that he has a password."
• negation as failure, e.g. "It is not provable that somebody enters a password."
• macros, e.g. "Proposition P: A card is correct. A clerk believes that P."
• passive voice, e.g. "John sees Mary." is treated equivalently to "Mary is seen by John." on the DRS level. Note that the subject must be explicitly mentioned in the by-phrase.
• numbers and strings as objects, e.g. "John's address is "Paris".", "The temperature is -2 in Paris in the winter."

Some ACE 4 constructs have been generalised in ACE 5, for instance:

• indefinite pronouns (somebody, everything, no one, ...) can have variables in apposition, e.g. "Everybody who loves somebody X likes X."
• proper names and variables can have relative phrases, e.g. "There is a man X. X who eats sees John who drinks."
• small numbers (0, 1, ..., 12) can be written as words (null, one, ..., twelve), e.g. "2 dogs see three cats."

For the details, read the updated ACE documentation and the updated DRS manual.

ACE 5 is backwards compatible with ACE 4, with a few exceptions:

• ACE 5 doesn't allow multi-word units where the words are separated by spaces. I.e. "No persona non grata has a valid zip code." must be rephrased using hyphens as "No persona-non-grata has a valid zip-code."
• Constructions "there is" + definite noun phrase are excluded in ACE 5. This means that sentences like "There is John.", "There is the man" and "There is X." are not valid ACE sentences anymore.
• Possessive constructions are now always interpreted as existential, i.e. "John's card" is "a card of John", "the man's card" is "a card of the man", etc.
• Possessive specifiers cannot be conjoined anymore, e.g. "John and Mary's dog barks." and "His and her dog barks." are not allowed in ACE anymore.

The new features are available as part of the APE webinterface and of the ACE webservice. Note that the URL of the webservice has changed. The new URL is http://www.ifi.unizh.ch/cgi-bin/attempto/ape/apews.cgi. The old URL will stay around for a while.

In addition to the support for the language extensions, the APE webservice offers an alternative way to paraphrase if-then sentences as every-sentences. E.g. "If a person owns a house then the house has at least 2 windows." is paraphrased as "Every house that is owned by a person has at least 2 windows.". There are also updates and bugfixes to the support for converting ACE texts into OWL DL. Read the related documentation for more.

Finally, we introduce ACERules --- an implementation of Grosof's Courteous Logic Programs --- that provide prioritised rules in ACE. A webinterface and a webservice for ACERules are in preparation.

As always, we welcome any feedback concerning ACE and its tools.

We are preparing version 5 of ACE that will --- among other features --- offer that-subordination, modality, negation as failure and the passive voice. Here is an example sentence showing some of the new features.

If there is some money and a customer cannot transmit the money and it is not provable that he does not transmit the money then the bank must assume that the money is transmitted by the customer.

This new functionality is currently being implemented, and will be soon available via ACE's web-interface and web-service. Wait for further news.

The biggest change in the new version of APE is the simplification of the DRS representation of nouns. See the updated DRS Report for the specification of the new format.

Also, the interface to the webservice has been made a lot simpler. See the updated APE Webservice Specification.

APE has also a preliminary support for OWL DL output, which means that ACE can now be used to write formal ontologies which are compatible with reasoners like Racer and Pellet, and tools like Protege and SWOOP. Documentation of this feature will be available soon.

The paraphrase feature has been extended to cover collective forms of plurals.

APE can now guess the syntactic properties of unknown words. In many cases this is a good alternative to the user lexicon or word prefixes.

The user lexicon now supports user-defined word types. See the updated ACE Lexicon specification.

And finally, APE has less bugs and is much faster than before.

This week we will be launching a new version of the APE webservice. As the new version introduces major changes to the DRS representation of ACE texts, we have backed up the old version of the webservice and its related documentation. Note that the backup will not stay available permanently. We ask you to update your applications to use the new DRS format.

The old webservice will stay available at: http://www.ifi.unizh.ch/cgi-bin/kalju/oldape/apews.cgi.

The old documentation and the webclient will stay available at: http://www.ifi.unizh.ch/attempto/oldtools/.