Section | Change |
Intervention | Please be sure to check what is and what isn't an Intervention. In particular "seeing" a patient is not an intervention, neither do we want general, non-specific words like "treated" to be annotated as Interventions. |
Signals must always modify something | Signals (Negation, Laterality, and Sub-location) must always modify something. They cannot stand on their own. In the case of Negation, this is a condition. For the other two, usually a locus. Please use Step 3 of the annotation recipe when you mark signals, add their "modifies" relation. Do not wait until you are annotating other relations to do this. |
Annotating relationships | Please read the general guidelines on annotating relationships. In particular, please try to only annotate those relationships that the text is telling you about. Often, such relationships are clearly stated. Sometimes, the text is saying something, but it needs some clinical knowledge interpret this and to decide on the relationship. This should not mean, however, that you try to deduce every single relationship between every single entity, regardless of whether the text is saying something about it. We are only interested in what the text is telling us. Please read the general guideline on this. |
Recipe: co-reference | Annotators are asked to take particular care with co-reference. Please look for co-reference as a separate step, before dealing with other relationships. The guidelines recipe has been updated to include this extra step. |
Negation values, Laterality values | Please make sure that you fill in the values for Laterality (right, left or bilateral) and Negation (absent or uncertain) |
Sub-location | Several more stock phrase examples have been added for Sub-location |
Co-reference of sets | Coreference annotations should not be created between things and sets that they are members of. See guideline for examples. |
General guidelines: sets | Note about annotating sets of things added. |
Drug or device: general drug classes | Note about coreference of general drug class terms added. |
Section | Change |
Complex loci | Annotating complex loci and sub-locations can be confusing. A tabulated recipe with examples has been added to help clarify this. Please try to use this recipe when faced with complex loci and sb-locations |
Negation | Note that "?" (meaning query) is a valid negation signal, with value uncertain. Added example to guidelines |
Result in histopathology | There are various ways of expressing positive histopathology stain and other test results, as well as the word "positive". Examples added. |
Condition | "Abnormality" is a valid Condition. See the example added in the Conditions section, sub-section "General terms for problems and diseases" |
Condition, Condition in histopathology | Cellular processes such as "transformation" and "prolifeation" should be marked as Conditions. See the relevant conditon section, and also Histopathology reports. |
Histopathology reports: Locus | Added further examples of cellular and tissue loci. |
Sublocation modifies relationship | Note that sub-location may modify more than one Locus. This is not currently possible in the Knowtator tool, but will be added shortly |
Signals: modifying entities | Signals modify nearby entities in the same phrase. They are often miss-annotated. The general guidelines section on signals has been expanded to discuss this |
Sub-location | Sub-location is about general language area words such as "upper" and "inner" that modify anatomy, and should not be used to annotate other anatomical terms. Further examples added to the relevant Sub-location section. |
Coreference | Please note that this section has been rewritten (and hopefully clarified) since Revision 1.43. This especially affects those annotators working prior to 16 February. Please re-read this section, if you have not read it recently. |
Negation | Please annotate the entire negating phrase. See the example in this section. Please note that there has been a small clarification of this section since Revision 1.43. This especially affects those annotators working prior to 16 February. Please re-read this section, if you have not read it recently. |
Revision | Date | Description |
Revision 1.43 | 08 Feb 2007 - 09:42 | Last revision prior to splitting across several files |
Revision 1.41 | 06 Feb 2007 - 17:40 | Changes in response to debug run 3. Version sent out for debug run 4. Also used for Sheffield internal debug run. |
Revision 1.38 | 19 Jan 2007 - 14:08 | A few incidental changes. "Developing the corpus" section removed: now documented elsewhere. |
Revision 1.31 | 02 Jan 2007 - 11:59 | Further changes made in response to debug iteration 2: further comments from annotators |
Revision 1.30 | 30 Dec 2006 - 17:31 | Changes made in response to debug iteration 2, difference analysis and initial comments. Version sent to annotators for 3rd iteration. |
Revision 1.26 | 18 Dec 2006 - 18:08 | Small number of outstanding issues from dry run error analysis, that were not caught in previous discussions and comments. No substantial changes, mainly the addition of examples to clarify. |
Revision 1.24 | 18 Dec 2006 - 10:57 | Issues from dry run comments incorporated. Version sent to annotators for second iteration. |
Revision 1.22 | 15 Dec 2006 - 15:50 | Decisions and discussions from December 2006 consortium meeting incorporated. Many draft comment boxes that related to these discussions also removed. |
Revision 1.18 | 07 Dec 2006 - 16:45 | Version circulated prior to December 2006 consortium meeting. Minor changes plus some introductory paragraphs. |
Revision 1.14 | 01 Dec 2006 - 16:49 | Version sent to Manchester for annotation dry run. All initial Sheffield and Manchester comments incorporated. |
Revision 1.11 | 22 Nov 2006 - 21:32 | Draft prior to adding in initial Sheffield and Manchester comments |