Linguistic Analysis Tools for Thesis Writing

Extract

[...] Finally, UDPipe highlights the syntactic relationships between words, creating dependency trees like the one below. This tool has several interests in writing a thesis in linguistics: for example, it will be very useful for analyzing sets of texts and identifying repetitive syntactic or morphological phenomena, regardless of the language in which the texts are written. For example, UDPipe is capable of identifying all the subordinate clauses of a text, identifying all the 'who' or 'what' while recognizing the function of each clause. [...]

[...] Talismane It also has the functionality to analyze texts, even the most complex and unpublished ones, syntactically. These features have the advantage, in the context of a thesis or a report in linguistics, of performing very fine and in-depth text analyses, and this on sometimes non-standard texts or even on oral transcriptions. Students also gain a significant amount of time in their writing as Talismane annotates text corpora very quickly and the results obtained are reliable. In addition, this tool is a significant aid for students as it treats all questions of syntax and morphology of the language in a thorough and specific way (agreement of the past participle, adapted use of the subjunctive . [...]

[...] Collaborative Annotation Tool Other tools have been developed to annotate texts collaboratively. This is the case of the platform developed by UKP Lab (Ubiquitous Knowledge Processing Lab) of the Technische Universität Darmstadt (Germany): INCEpTION3. This tool promotes the creation and enrichment of texts through an open-source platform that allows for manual and assisted annotations. Its features include manual linguistic annotation, active learning through annotation suggestions made by the platform, and the creation of new models based on those present in the platform (allowing for training in manipulating NLP models). [...]

[...] The application identifies more stylistic errors, such as repetitions, long, heavy or incorrect sentences? and proposes examples of improvement. In the context of a thesis, this tool helps to improve writing, by improving spelling and grammar and seeking to make it fluid and pleasant to read for an examiner. For FLE students, having explanations about errors facilitates learning by clarifying errors and designing adapted exercises to help overcome them, thus individualizing student learning. For example, if a student writes « She went swimming at the pool », Scribens will detect the agreement error and propose the correction 'She went swimming at the pool' by adding the explanation 'The verb 'aller' is in the infinitive form, but it should be in the past participle form 'allée' to agree with the subject 'elle'.'agreement of the past participle with the feminine subject when the auxiliary is 'to be' and will then build training exercises on the agreement of the past participle. [...]

[...] - URL of Prolexis : https://www.diagonal.fr/prolexis.html 5. Natural Language Processing (NLP) Platform Then, SpaCy 7SpaCy is a Python-based tool belonging to the German company Explosion AI, which specializes in NLP. It is a library with integrated visualizations designed to improve users' professional skills. Its features are centered around morphosyntactic tagging of words, syntactic analysis, entity recognition, linguistic pattern detection, all in a multi-lingual mode: in fact, the tool is usable in French, English, Spanish, and many other languages. The interest of this tool can be multiple. [...]