SC740 SEMINAR WEEK 7

TOPIC : A Knowledge-Based Approach for the Conflation of Vector

Mapping Data

By Dr. Maria Cobb

University Of Southern Mississippi

Summary and Comments : By Lee Emmanwori

A) Dr. Cobb introduced the concept of A knowledge-Based Approach for the Conflation of Vector Mapping Data. She explained map conflation with a diagram that looks like Map1---->A single better map----->Map2 a more accurate map.

B) Dr. Cobb pointed out some phases of conflation as:

1.) Feature Matching

2.) Feature Deconflation

3.) Positional Alignment-overlaying two digital maps

C) Dr. Cobb spent most part of the time explaining Feature Matching Process of the conflation phase. She outlined the process as involving--

1) Pre-processing step to unify logical feature parts segmented during collection or production.

2) Hierarchical rule with matching criteria based on :

- distance measure

- cartographic representation such as point, line, or area

- FACC feature and attribute codes

- neighborhood geometry

- topology such as bridge crossing road

3) User interaction--user to decide if the features are the same or not

D) She gave more examples of matching value of features F1 and F1 based on overall evidence F1 is the same as F2 using the formula

Vmatch = Sum(s(a1i a2j))*H12

--------------------------

Wes

Wes - is weighting factor returned by expert system

H12 - is normalization factor-count of matching attributes

s(a1i a2j) is the similarity of code values a1i, a2j

s = +1 values are identical

0<s<1 values similar to some degree

s = 0 neutral

-1<s<0 values dissimilar to some degree

E) Dr. Cobb pointed out some potential problems as

- missing information

- reliability of information present