Cassandre

From Hypertopic

1 Features
2 Aims
3 Inside look
4 Development credits
5 Cassandre Home Page
6 Download Cassandre

Features

Cassandre is designed for:

Exhibiting peculiarities (of texts),
Identifying common features (of texts),
Analysing texts,
Interpreting texts,
Confronting alternate interpretations.

Aims

The platform aims to provide scientists with tools to annotate, analyse and interpret textual materials (such as interview transcripts, discussion forums or press articles). Analyses can be constructed, shared, compared among users.

Analysis are conceived as relations (made by the user) between materials and conclusions. Laying in the platform, these relations are explicit, exhibited, available and accountable for scientific discussions. Resting on such explicit relations, conclusions becomes stronger.

Both top down (from theory to material) and bottom up (from material to theory) approaches are allowed. According to user's methodological options, annotations can consist in content analysis, thematic analysis, discourse analysis, category-bound analysis, interpretive analysis.

Inside look

Sample use of Cassandre with Porphyry and Argos

Cassandre complies to the Hypertopic model but has a few specificities.

First, as for other Hypertopic tools, topics name and relations are decided by the user. Anyway, “leaf” topics have a special behaviour. When a topic has no child, the software consider its name as a keyword: any fragment that comprise this string is automagically bound to the leaf topic. These search patterns include regular expression capabilities.

Secondly, topic networks inside Cassandre are trees. Relations are always “topic-inclusions”. As described on Porphyry's page, this relation type can be used to express several meanings... including collection of “leaf” topics.

Thirdly, Cassandre generates standards attributes by parsing texts. Inspired by lexicometrics and natural language processing, attributes comprise frequencies, collocations, hapaxes and specificities. These measures should be used as clues for the user to construct an analysis relevant to his/her research questions.

Thus, the Hypertopic platform provides two approaches to qualitative analysis of textual material. On the one hand, Cassandre's design furnishes a keywords/patterns semiautomatic approach. On the other hand, Argos server offers a user-coded segments approach, which can be conceived as the traditional widespread QDA approach. The platform allows to combine both these approaches.