Text contents of 3 tables some how related to each other But there is no linear direct relationship between them e.g First row of table 1 is not related to first row of table 2 and table 3 although they sharing same numeric and letter key but may be there is relationship between First row of table 1, second row of table 2 and first row of table 3. so table 1 is related to both table 2 and table 3 via sharing common numeric key and letter key but not in orderly way.
You can consider Table 1 as questions table each row is individual question and Table 2 is Responses from one group of people and table 2 is response from other group of people. But in collected responses some people did not reply the answers orderly way . for example they Gave answer of first question then skip the answer of second, third and wrote answer of fourth one. all of data was collected on paper. so they just place bullet points in front of answers without specify the the answer belong to what number of question.
I have to figure out some insight from data, so what kind of data analysis or approach should I use to analysis this unstructured data.
I am already using Power BI for creating word cloud to visualize the data. What kind of other machine learning approach I can use to figure some insights from this data