Automatic detection of variable data in Web document: Graphical representation on demand

Graphics on demand: the automatic data visualization on the WEB

“…The last step consists in identifying the variable data: the "Main Variable" that depends on the context and the others according to which the Main Variable varies. We have developed an algorithm [9] allowing to extract the different variables and to "format" them in a structure such as a table. The algorithm builds a matrix M with n rows (n is the number of sentences in the document) and three columns containing respectively nouns, verbs and complements, which are obtained from each sentence.…”

“…The process of identification of significant variable data takes place in three main steps [9]: (i.) the grouping of the named entities in classes of categories, (ii.)…”

“…Thus, in a second step, the identification of the different complex nominal groups found in the document is required in order to have a more accurate context. An algorithm allowing the identification of the context has been implemented and experimented [9].…”

“… The Identification of the main topics [9] of a document helps us to identify the most important textual units of a document. These units are formed by key-words that we refer to in this document as "key-terms candidates and are the different noun phrases in a document.…”

“…However important information is missing: the date. The algorithm presented in [9] determines the different complex nominal groups and identifies the final context: "The French presidential election of 2017".…”