MHDBDB-Help: Text Analysis
The information system of MHDBDB retrieves meaningful responses to almost any linguistic and semantic question asked about the textbase. For working analytically with individual texts, text groups or the entire text archive the module Text analysis should be used.
It offers the following basic search possibilities:
Such searches will apply to the entire vocabulary, in other words, includes words that have not yet been lemmatized.
- for individual words or character strings
- for several different words at the same time
- for several words in sequence
- for words before or after punctuation marks or caesuras
- for words in various positions within a line (important, e.g. for metrical analyses)
- for simple to more complex word combinations within a given context.
The search results appear in form of a table. Each individual text is represented in the table by its short letter code, below which appear the number of references retrieved from the text for each word. By clicking on these reference numbers you can call up the text for the references (see below).
The search can be applied to:
You arrive at your selection by using the appropriate functions and/or highlighting the individual texts or groups.
- an individual text,
- a number of texts selected by the user,
- pre-selected groups according to author or text-type
Before searching for two or more words within a given context (context search) you have to define or choose the size of the context by selecting the respective range of words or lines within which the respective words should co-occur. Similarly, you may determine the size of your resulting text references by selecting the number of lines before and after the key words found in the text(s).
Each search command has to be stated according to certain syntactical rules. You will find detailed descriptions and examples on that in the help documents (see: General Help
and Text query help).
The most important syntactical rules are the following :
The search results are at first always rendered in form of a table consisting of the short acronyms for individual texts, the individual words and the numbers for the retrieved references. The basic words (lemmas) (e.g. klagen;
vrouwe) are highlighted in yellow; the variants assigned to these lemmas (e.g. chlagn,
appear without highlighting before the basic page background. Words that have not yet been assigned to a lemma always appear at the end of the table.
- In general:
- no spaces between search terms and function symbols;
- searches for a specific string of characters:
- must be preceded by „$“;
- searches for a lemma and all its variants
- must be preceded by „@“;
- searches for words in sequence:
- the words must be separated by a comma;
- searches for a word that also meets certain grammatical or semantic criteria or appears in a certain position within a text-line:
- the criteria must be joined by a „&“;
- searches for words within a given context:
- the search terms must be joined by a „+“;
- simultaneous searches for different words or linguistic or semantic characteristics:
- the terms must be joined by a „|“ (= oder).
By clicking on a reference number you will arrive at the actual text references. The retrieved words are highlighted in red within the context of each text reference.
Not only can you search for individual words but for entire word fields belonging to a given conceptual area.
For such searches according to specific conceptual catgegories (represented by numerical codes within the system, see What is it? (Dictionary)) you have several options that also have to be formulated according to the basic syntactical rules given above. For example:
Tristan und Isold (Gottfried von Straßburg), Zeile 116 - 124
- You may search for all words and/or their variants in MHDBDB, which relate to the conceptual area „DEATH“ fallen;
- or which relate to the conceptual area „LOVE“;
- you may also search for words which belong to both conceptual areas at the same time (attention: you must combine the search terms using &, e.g.:
21104000&22704010, result: minnetôt, tôtviendinne
- you may also search for text references where, for instance, within a given context, words from the conceptual area „DEATH“ co-occur with words from the conceptual area „LOVE“ (attention: here the search terms have to be separated by the symbol +, e.g.: 21104000+2704010. Result
(context: 3 lines; lines before and after the keywords: 3 each):
||daz übel daz tuot so herzewol,
||daz es kein edele herze enbirt,
||sit ez hie von geherzet wirt.
||ich weiz ez warez alse den tot
||und erkennez bi der selben not:
||der edele senedaere
||der minnet senediu maere.
||von diu swer seneder maere ger,
||dern var niht verrer danne her;
Again, the conceptual categories are represented by numerical codes within MHDBDB. Find more detailed explanations on searches with conceptual categories in What is it? (Dictionary) and in
General Help - Conceptual Categories.
Back to Help Contents