All videos are in german language.
Please note the changes in ANNIS 4 listed at the end of this page.
1. Das ReN - Projektwebseite und Korpusstruktur (ReN - Website and Structure of the Corpus)
2. Zugang zum Korpus (Access to the Corpus)
3. ANNIS-Oberfläche (The ANNIS Interface)
4. Transkription und Annotation (Transcription and Annotation)
5. Suchen im ReN in ANNIS (Searching in the ReN with ANNIS)
6. Export und Frequenzanalyse (Export and Frequency Analysis)
7. Beispielsuchabfragen im ReN in ANNIS (Example Search Queries in the ReN with ANNIS)
Changes in ANNIS 4
The view of the corpus list below the search query window has changed slightly. Under Visible, you can no longer choose between the annotated part of the ReN (ReN-anno) and the transcribed part of the ReN (ReN-trans). Instead, only "All" is listed, i. e. the entire corpus of ReN-anno plus ReN-trans. To activate one or more sub-corpora within the displayed corpus list, simply click on the box to the left of the corresponding sub-corpus to activate it. By clicking on the box at the top left of the line Corpus, all the sub-corpora displayed in the list can be activated. To systematically select several sub-corpora, e.g. of the same language area, the filter function can also be used. If, for example, you enter the entry ostelbisch under Filter, only all sub-corpora from the East Elbian area will appear in the list; by clicking on the top box, all these sub-corpora can be activated at the same time. In this way, it is also possible to limit the list to the sub-corpora "ReN_anno" and "ReN_trans", e. g. the entry ReN_anno can be made in Filter, and only the annotated sub-corpora will appear in the corpus list.
View the metadata for a specific text:
To view the metadata for the text of a specific subcorpus in the corpus list, click on the document browser icon to the left of the name of the subcorpus in the corpus list, which looks like a page of paper (to the right of the info button). Then the window opens in which the texts contained in the subcorpus are listed. To the right of the selection of the various transcription views is the column info. If you click on the info button in the line of the corresponding text, a small window opens with the metadata of this text. This window can also be enlarged by dragging it with the mouse.
Search including metadata:
To include metadata in the search, the part-of operator @* must be used. With the search query pos="PPER" @* topic_ReN="R", for example, a search is made for all personal pronouns in legal texts (topic_ReN = field of writing, R = law).
If several metadata are to be included in the search, a reference to the annotation searched for must be created for each of the metadata using the node number, for example, for a search for all personal pronouns in North Low Saxon legal texts:
pos="PPER" & #1 @* topic_ReN="R" & #1 @* language-area=/nordniedersaechsisch/
Extension of the search options:
ANNIS 4 enables a number of new searches, e.g.
- negation of operators with existence (from 4.4.0), see https://github.com/korpling/graphANNIS/issues/186
- negation without an existing node (from 4.5.0), see http://korpling.github.io/ANNIS/4.5/user-guide/aql/negation.html
Results of a search query:
Basically, the search queries in ANNIS 4 run much faster than before in ANNIS 3, and very complex queries are also possible. In the case of a query in a large number of sub-corpora or a very complex query, it can happen that the hits are already listed in the results window on the right, but the number of matches does not appear below the query window. In these cases, simply click on the search button again and the number of matches and documents will appear.
Metadata is no longer included via the button Select in the input window. The reason is that frequency annotations in ANNIS 4 are treated like normal annotations in AQL. If you want to perform a frequency analysis including a metadatum, this metadatum must already be attached to the search query using the expression @*, e.g. you can use pos="NE" @* abbr_ddd in order to search for all proper nouns first, then when opening the frequency analysis in the new window, the metadatum abbr_ddd (= sigle of the text) appears in the list of nodes besides the searched annotation level pos. To filter the hits for the proper nouns according to the texts, the node pos must be deleted and then the frequency analysis is started.
There is now only one CSV exporter. This corresponds to the CSV-Multi-Tok-Exporter in ANNIS 3.