An Artificial Intelligence Approach to Modeling in Social Science

Computer Science has contributed to social sciences since decades ago: connecting people that build virtual communities where the interactions can be investigated, developing tools for statistically analytics, designing models allow analysis and simulation of most diverse types, among many others. In this article, we describe an artificial neural network model a theoretical framework risk, housing, health problematic, called DRVS (Diagnostic methodology risk determination urban housing health), which uses holistic approach community environmental health. The also exposes digital clinic history families communities, developed support acquisition necessary data. This software advantages transference application in different locations it constitutes expert system local indexes supports quantitative validation process underlying theory. On other hand, as intelligence techniques, constraints: unlike explicit logic inferences, networks work «black boxes», not explaining how they got result; have strong dependency representativeness training data introducing new knowledge may improve their results performance is difficult (new data, addition or remotion determining factors model, weighting factors, etc.). article shows some techniques ideas on deal with identified constraints.