Skip to content
Dirk Roorda edited this page Oct 16, 2017 · 6 revisions

Data module

This project delivers a new data module with verbal valence data, to be used alongside the core ETCBC data set of the Hebrew Bible bhsa.

The new module resides in this repository, at this location.

Feature description

Here is a description of the new data features.

name workflow values description
cfunction correction undefined, empty or values as in function corrected phrase function, only defined for phrases that occur in a correction sheet, empty if no correction has been entered in that sheet, otherwise a valid phrase function code
f_correction correction undefined, empty, or y only defined for phrases that occur in a correction sheet, empty if no correction has been entered in that sheet, y otherwise
s_manual enrichment undefined, empty, or y only defined for phrases that occur in a enrichment sheet, empty if no enrichment has been entered in that sheet, y otherwise
predication enrichment see table predication below main classification of the verbal function
valence enrichment see table valence below main classification of the verbal valence
grammatical enrichment see table grammatical below indicates that the constituent is subject or object, with a fine distinction between kinds of objects
original enrichment see table grammatical below as grammatical, but these are default values, before more sophisticated enrichment logic has been applied
lexical enrichment see table lexical below indicates that the constituent carries certain semantic characteristics based on lexical information
semantic enrichment see table semantic below indicates that the constituent carries certain semantic characteristics, based on other than lexical information

Enrichment values

Predication

value description
NA not applicable
regular describes action or state
copula links subject to a predicate

Valence

value description
core core constituent, such as subject or direct object
complement constituent that complements the verb
adjunct constituent that is loosely coupled to the verb

Grammatical

value description
NA not applicable
subject subject
principal_direct_object main direct object among multiple direct objects
direct_object direct object in case there is only one
NP_direct_object direct object in the form of a noun phrase, only if there are multiple direct objects
indirect_object indirect object, not to be confused with benefactive adjuncts
L_object complement starting with preposition L, having certain characteristics of a direct object; only if there are more direct objects
K_object complement starting with preposition K, having certain characteristics of a direct object; only if there are more direct objects
infinitive_object complement in the form of a clause with an infinitive predicate; only if there are more direct objects

Lexical

value description
location locative complement/adjunct
time temporal complement/adjunct

Semantic

value description
benefactive complement/adjunct that "benefits" from the action, not to be confused with an indirect object
location locative complement/adjunct
time temporal complement/adjunct
instrument instrumental complement/adjunct
manner complement/adjunct of manner

Annotation set

We have bulk-uploaded a set of annotations to the website SHEBANQ. This set is a CSV file in this repo.

It has a row for each assignment of a sense label to a verb. The fields of the rows adhere to a format specified by SHEBANQ.

The text of each annotation specifies

  • the verb occurrence;
  • the sense label of the verb in this context;
  • a link to the flowchart of this verb if there is one, otherwise a link to the legend of sense labels;
  • a list with the names and values of the key constituents on which the choice of sense label has been based.

All constituents are given as a number referring to a phrase or clause in the context and a sequence of glosses in English.

Flowcharts

  • BR> - ברא - create
  • DBQ - דבק - cling
  • NTN - נתן - give
  • <FH - עשׂה - make
  • QR> - קרא - call
  • CJT - שׁית - set
  • FJM - שׂים - put
  • ZQN - זקן - be old

Clone this wiki locally