We continue the discussion on Text Knowledge Mining. We now look at what KM does not mean. In other words, we look at what people are saying about Knowledge mining. For example, knowledge mining is employed to specify that the intermediate form employed contains knowledge richer than simple data structures though inductive data mining techniques are used on the intermediate forms and the final patterns contain only those that are interesting to the user. Another example is when knowledge mining is employed to indicate that the background knowledge and user knowledge is incorporated in the mining process in order to ensure that the intermediate form and the final patterns contain only those concepts interesting for the user. In yet another example, knowledge mining refers to using background knowledge to evaluate the novel and interesting patterns after an inductive process.
Sometimes another term deductive mining is used but it is vague and even this is different from the proposal made so far. Deductive mining is used for a group of text mining techniques  where the better known example is said to be information extraction. This is referred to as the mapping of natural language texts into predefined, structured representation, or templates, which when filled represent an extract of key information from the original text or alternatively as the process of filling the fields and records of a database from unstructured text. This implies that no new knowledge is found.  Therefore this falls in the deductive category based on the starting point but the difference is that the deductive inference performed is translating from one representation to the other. One specific application automatic augmentation of an ontology relations is a good example of information extraction. TKM can use such techniques.
Some have added that deductive data mining be performed by adding deductive capabilities to mining tools in order to improve the assessment of the knowledge obtained.  The knowledge is supposed to be found mathematically and taken into account hidden data assumptions.
A deductive data mining paradigm is also proposed where the term refers to the fact that the discovery process is performed on the result of the user queries that limit the possibility to work on corrupt data. However in these cases, the generation of new knowledge is purely inductive.
Having discussed what TKM is different from, we now see what TKM is related to. Some have alluded to non-novel investigation like information retrieval, semi-novel investigation like standard text mining and KDD for pattern discovery and novel investigation or knowledge creation like Intelligent Text Mining which implies the interactions between users and text mining tools and/or artificial intelligence techniques. In the last category here, there is no indication about the kind of inference used to get the new knowledge.
Some advanced text mining techniques that can be considered TKM are related to literature based discovery where these techniques provide possible relations between concepts appearing in the literature about a specific topic although some authors beg to differ from discovery and mention assisting human experts in formulating new hypothesis based on an interactive process.
#codingexercise
Double GetAlternateOddNumberRangeMax()(Double [] A)
{
if (A == null) return 0;
Return A.AlternateOddNumberRangeMax();
}
#codingexercise

Double GetAlternateOddNumberRangeAvg()(Double [] A)

{

if (A == null) return 0;

Return A.AlternateOddNumberRangeAvg();

}

#codingexercise

Double GetAlternateOddNumberRangeSum()(Double [] A)

{

if (A == null) return 0;

Return A.AlternateOddNumberRangeSum();

}