Yes, first define the predicative format of your ontology; normally it would be a register of entities and a register of properties that may be their predicates. Second, provide a close grammatical analysis of the sentences of your text. Then squeeze out the predicative transformations you can get, and place them in the ontology. Range ontologies according to semantic domains — otherwise you get chaos.

Thanks, Per Aage Brandt,

Is there any automated tool that can help me?

I recommend that you program it by yourself.

ok. Thanks

From a social realist perspective (Archer 1995) texts are representations of culture, structure and agency. Each text we read, especially critical analysis, usually makes explicit or implicit references to structures (schools, curricula, knowledge), culture (disciplinary, community, institutional) and agency (students, peers).

Eg: "The students experience alienation in the classroom as they could not understand the language fully and relate to the teachers' perspectives".

Students and teachers = agents; classroom = structure; culture = language & teachers' perspective.

Oscar Oliver Eybers Thanks. agreed with the shared text.

Or: agents and experiencers; action or experience frames; and content of action or experience. Ontological grids like the proposed may be useful.

Per Aage Brandt manual or automatic? any tool for semi-automatic or automatic construction of domain ontology?

You may check out the framenet devices, close to George Lakoff. Or ask him directly — with greetings from me. :)

Per Aage Brandt Thanks

Abdul Sattar Did you find a solution? I worked on this issue and my idea was to convert texts to the graph (I am not the first one who has such idea) where conversion algorithm can use in the first step chunking to construct nodes of the ontology graph and the arcs can correspond to keywords (verbs) found between the entities. In the second step, chunks can become entities and arc relations in the ontology graph. You can find this useful https://en.wikipedia.org/wiki/Ontology_learning.