it is true that ethnographic or anthropological studies can aid our understanding of kinship, but which architecture characterizations and spatial distribution pattern can be of reference to specific type of kinship?
Three terms in the question formulation that need clarification, before an answer can be given: infer, interpret, and remains. Starting from the last one first: are you asking how archaeologists may guess at kinship systems by examining architectural elements left behind by people who are no longer there? If yes, then the short answer is: it is nearly impossible and quite hazardous to try such an undertaking. Even in the presence of some documentary evidence (letters, wills, etc.) in conjunction with architectural remains from the past, any guesses/hypotheses would in most cases be unreliable.
Now to "infer" and "interpret". If, on the other hand, the context is one in which the people whose architecture you are studying are still around, then the answer is: ask them. Compared to drawing inferences from the forms, materials, configurations, etc. of buildings, it is much safer to solicit responses to the questions that interest you, and then interpret these responses. The reason for this is simple: inferences re. social/community/interpersonal practices must by necessity presuppose some 'theory' about the relationships between buildings and humans, e.g., that buildings "express" or "reflect" the values, practices, etc. of those who build/use them. Attractive as such theories may be (at first sight), they are almost invariably misleading. Think of all the buildings or neighbourhoods that you immensely dislike and would love to see them gone. As a member of the society that produced these things, do they "reflect" your values?
this article may help. You will still need ethnographic records of local kinship, and suppose such kinship system practiced certain veneration or rituals.
Mantha, Alexis
2009 Territoriality, Social Boundaries and Ancestor Veneration in the Central Andes of Peru. Journal of Anthropological Archaeology 28(2):158-176.