Algorithm for Exposing Knowledge

Communication with people in the Arabic world had taught me to harness my ever-thrilled attitude of sharing knowledge, expressing opinions, and expecting some interaction and response from others . I have learnt the following algorithm:

If OtherParty is interested, and have close level then expose your knowledge/opinion/info.

else post publicly blackboard method) and interested ones can search and find it

end

Notice that it has no rule/instruction for taking explicit initiatives for approaching people!

Personal Thoughts on Database Design Instructions

A typical scenario for a student in a database design course would go like this: the conceptual modeling is first taught, then its transformation into a logical model, then comes the normalization process.While this methodology is systematic, with its well-established rules of thumb and standardization procedures, but when provided to the student in this style; i.e. as rules of thumb to follow, it can miss a crucial component: the component that would associate the procedural rules with the ‘common-sense’.I will draw a metaphor from the artificial intelligence programming field, specifically the agent framework. The simplest kind of agent is a non-cognitive agent. Rather than directing activity based on reasoning, a non-cognitive agent simply reacts based on built-in conditioned reflexes in the form of: if condition1 occurs then action1, if condition2 then action2,..etc. Whereas a more complex kind of the cognitive agent, with mental representations of the world and likes and dislikes. One very famous architecture for such cognitive agents is the Belief-Desire-Intention (BDI) agent that is capable of reasoning and generating the appropriate plan of actions. I would say that the reflexive agent can be well-instructed to perform the normalization checking procedures if provided with built-in rules of thumbs of normalization such as: if transitive dependency exists in a relation (table) then remove the attributes with transitive dependency into another table. Whereas the cognitive agent would be designed with mental states and preferences such as: beliefs (to represent the business rules), desires (to represent preferences of domain-separable and duplicate-free relations), and Intentions (that would represent the integrity constraints and information-accessibility minimized). Thus, in this case the BDI DB can be read as “Business rules-Domain separable/duplicate free- Integrity satisfaction/information-accessibility minimization” DB. The normalizations procedure would be the outcome of reasoning/planning with these mental states in mind in this case, rather than being the ends by itself. This can be illustrated by the following example.

propertyNo

pAddress

staffNo

sName

iDate

iTime

carReg

comments

 With the business rules of:·         a staff is assigned a car for one-day inspection, ·         a property cannot be inspected more than once in a dayApplying the domain separable concept would suggest that Property and Staff attributes each be in a separate relation (table). While applying the duplicate-free preference would suggest that the carReg is required to be separated (because for the same date several properties can be visited with the same carReg number) along with staffNo and iDate for the sake of information-accessibility minimization, that is a staffNo and iDate are the minimum information required for the retrieval of carReg.  That is not to say that the classical systematic and methodical normalization are to be replaced, on the contrary, that was a reminder of the necessity of always associating the rules with a higher mental attributes of objectives (intentions), domain requirements (beliefs) and conditions to satisfy (desires), especially when these rules are to be performed procedurally by a cognitive entity. This way, the procedural rules will be always associated with and derived by common sense.