If I look, e.g. for ቅድሳት፡, I will get the correct result with the best match: the lemma ቅድሳት፡, but the equally pertinent ቅዱስ፡, which has a plural form ቅድሳት፡ will be placed very far away. This could be solved by adding an information about the root and sorting according to that.
the query should also search the strings which are possible derivate of its root.
If I look, e.g. for ቅድሳት፡, I will get the correct result with the best match: the lemma ቅድሳት፡, but the equally pertinent ቅዱስ፡, which has a plural form ቅድሳት፡ will be placed very far away. This could be solved by adding an information about the root and sorting according to that.
the query should also search the strings which are possible derivate of its root.