Logo SfSGermaNet

Database

The working copy of GermaNet is stored in a database. It contains all the GermaNet data. The GermaNet Editing Tool (GernEdiT) is used by the lexicographers to enter lexicographic data into the GermaNet database. For more information on GernEdiT, please refer to the following paper:


Verena Henrich and Erhard Hinrichs: The GermaNet Editing Tool. In Proceedings of the Seventh Conference on International Language Resources and Evaluation (LREC 2010), Valletta, Malta, pp. 2228-2235.
[Download paper: http://www.lrec-conf.org/proceedings/lrec2010/pdf/264_Paper.pdf]



Database Structure

The database is structured according to GermaNet, which means that there are tables to store synsets, lexical units, relations, etc. Figure 2 overviews the whole database structure. Each table is explained in detail in the below.


Entity-Relationship-Diagram
Figure 2: Overview of the database structure (click here to enlarge)


lex_unit_table

Each entry in this table represents a lexical unit with all its information.

  • id: unique identifier
  • synset_id: specifies the synset, to which this lexical unit belongs to
  • orth_form: the orthographical form of this lexical unit
  • orth_var: an orthographical variant
  • old_orth_form: the orthographical form, which specifies a difference in the old German orthography; this is only set, if there was a second allowed variant in the old orthography
  • old_orth_var: an orthographical variant, which specifies a differing variant, which was allowed in the old German orthography; this is only set, if it is not allowed anymore in the new orthography
  • named_entity: specifies whether this lexical unit is a named entity or not
  • artificial: specifies whether this lexical unit is used to represent an artificial node in the graph
  • style_marking: specifies whether the style of this lexical unit is marked
  • comment: a comment for this lexical unit


example_table

Each entry in this table represents an example that belongs to a lexical unit.

  • id: unique identifier
  • lex_unit_id: refers to the lexical unit, to which this example belongs to
  • text: the example itself
  • frame_type_id: specifies the frame type of this example; if this is empty, this example has no frame


frame_table

Each entry in this table represents a frame that belongs to a lexical unit.

  • id: unique identifier
  • lex_unit_id: refers to the lexical unit, to which this example belongs to
  • frame_type_id: specifies the frame type of this frame


frame_type_table

This table contains all possible frame types.

  • id: unique identifier
  • frame: the frame type itself, e.g. NN, NE, or NN.AN.Az


lex_rel_table

All lexical relations are stored in this table.

  • id: unique identifier
  • rel_type_id: specifies the type of lexical relation, see the description of lex_rel_type_table
  • from_lex_unit_id: specifies the lexical unit from which this lexical relation starts
  • to_lex_unit_id: specifies the lexical unit to which this lexical relation goes


lex_rel_type_table

Here, all types of lexical relations are stored.

  • id: unique identifier
  • name: the name of the lexical relation, e.g. antonymy or pertonymy; notice that synonymy does not appear in this table, because the synonymy relation can be determined by searching for all lexical units with the same synset_id
  • direction: specifies if this lexical relation is valid in one or both directions (i.e. one or both); revert means that the relation is in both directions, but in different ways
  • inverse: the name for the lexical relation in the inverse direction; especially interesting, if the direction field (dir) is specified as revert


synset_table

Each entry in this table represents a synset with all its information.

  • id: unique identifier
  • word_class_id: specifies the word class of this synset, e.g. Bewegung, Geist, etc.
  • word_category_id: specifies the word class of this synset, e.g. adj, nomen, or verben
  • paraphrase: a description of this synset
  • comment: a comment for this synset


word_class_table

This table stores all possible word classes.

  • id: unique identifier
  • word_class: the word class itself, e.g. Allgemein, Bewegung, Geist, etc.


word_category_table

This table stores all possible word categories.

  • id: unique identifier
  • word_category: the word class itself, e.g. adj, nomen, or verben


con_rel_table

This table contations all conceptual relations.

  • id: unique identifier
  • rel_type_id: specifies the type of conceptual relation, see the description of con_rel_type_table
  • from_synset_id: specifies the synset from which this conceptual relation starts
  • to_synset_id: specifies the synset to which this conceptual relation goes


con_rel_type_table

Each entry in this table specifies one type of conceptual relation.

  • id: unique identifier
  • name: the name of the conceptual relation, e.g. hyperonymy or meronymy
  • direction: specifies if this conceptual relation is valid in one or both directions (i.e. one or both); revert means that the relation is in both directions, but in different ways
  • inverse: the name for the conceptual relation in the inverse direction, e.g. hyponymy; especially interesting, if the direction field (dir) is specified as revert
  • transitive: specifies whether this conceptual relation is transitive


ili_table

This table contains the ILI Records which map GermaNet to WordNet

  • id: unique identifier
  • gn_lex_unit_id: lexical unit id in the GermaNet database
  • gn_old_sense: deprecated
  • ewn_relation:specifies the type of relation, see the description of ewn_rel_table
  • pwn_word: an English translation of the GermaNet lexical unit
  • pwn_word20_sense: a one digit hexadecimal integer that uniquely identifies a sense within a lexicographer file
  • pwn_word20_id: the id of the WordNet 2.0 synset matching the GermaNet lexical unit
  • pwn_word30_id: the id of the WordNet 3.0 synset matching the GermaNet lexical unit
  • source: describes whether the data originates from the EuroWordNet project or from the University of Tübingen
  • pwn_word20_paraphrase: definition from WordNet


ewn_rel_table

This table contains all possible cross-lingual relation types from EuroWordNet.

  • id: unique identifier
  • name: the name of the conceptual relation between GermaNet lexical units and WordNet synsets.


ili_pwn20_synonym_table

This table contains alternative translations of a GermaNet lexical unit.

  • id: unique identifier
  • ili_id: id of the ILI record
  • pwn20_synonym: alternative translation of a GermaNet lexical unit, taken from WordNet synsets


wkn_sense_table

This table contains links from GermaNet lexical units to Wiktionary paraphrases.

  • id: unique identifier
  • gn_lex_unit_id: lexical unit id in the GermaNet database
  • wkn_word_entry_id: id of the Wiktionary entry
  • wkn_sense_id: id of the Wikntionary sense, within an entry
  • wkn_gloss: the paraphrase of the GermaNet lexical unit as taken from Wiktionary
  • wkn_gloss_edited: indicates whether the previous column has been altered for GermaNet


compound_table

This table contains information about constituents comprising nominal compound lexical units.

  • id: unique identifier
  • comp_property: specifies a property of the compound. See compound_property_table
  • compound: orthographic form of the compound
  • comp_lex_unit_id: id of the corresponding lexical unit in lex_unit_table
  • mod_property: specifies a property of a compound modifier. See compound_property_table
  • mod1_category: specifies the category of the first compound modifier. See compound_category_table
  • modifier1: orthographic form of the first modifier
  • mod1_lex_unit_id: id of the corresponding lexical unit in lex_unit_table
  • mod2_category: specifies the category of the second compound modifier. See compound_category_table
  • modifier2: orthographic form of the second modifier
  • mod2_lex_unit_id: id of the corresponding lexical unit in lex_unit_table
  • head_property: specifies a property of the compound head. See compound_property_table
  • head: orthographic form of the compound head
  • head_lex_unit_id: id of the corresponding lexical unit in lex_unit_table


compound_property_table

This table contains compound properties, e.g. Abkürzung, Wortgruppe, Eigenname.

  • id: unique identifier
  • name: specifies a property of the compound or its constituents, e.g. Abkürzung, Wortgruppe, Eigenname


compound_category_table

This table contains compound modifier categories, e.g. Nomen, Präposition, Pronomen.

  • id: unique identifier
  • name: specifies a category of a compound modifier, e.g. Nomen, Präposition, Pronomen