Search

This paper presents knowledge representation structure and knowledge processing techniques implemented in ESSM (expert system for the design of structural members) for efficient knowledge processing. The knowledge in ESSM has been represented in the form of rules with their structure improved to make rule processing efficient. The concept of grouping the rules in control groups (CGs) relieves the Inference Engine of the tasks of identifying the rules pertaining to a particular design step and checking and transferring of design parameters. The CG hierarchy can be scanned by using various knowledge processing techniques, i.e. depth-first search, breadth-first search and best-first search. The rules within a CG can be processed by using forward chaining or backward chaining methods. These knowledge processing techniques have been implemented in ESSM in LISP. The paper discusses their suitability in processing the structural design knowledge.

A case of recurrent vesicogluteal fistula due to a spiculated sequestrum of the acetabulum, a sequela to treatment including excision of the head of the femur for infective arthritis of the right hip, is reported.

Efficacite du test d'hemagglutination indirecte pour doser les antigenes a differentes etapes de la production de vaccin contre l'encephalite japonaise prepare sur cerveau de souris

The reverse indirect haemagglutination (RIHA) test has been standardized for the assay of the viral antigen of purified Japanese encephalitis (JE) vaccine. Glutaraldehyde fixed sheep erythrocytes were sensitized with ammonium sulphate purified antibodies to JE vaccine raised in mice. The sensitivity of the erythrocytes fell to about one hundredth of the initial sensitivity in the first two days after preparation. After initial loss in sensitivity the stability of the cells became stabilized and the cells retained their titre for one year at 4-8 degrees C. The initial loss in sensitivity was not reduced by storing the cells at -70 degrees C, but after freeze drying the sensitized cells with a stabilizer one day after their preparation the cells retained their sensitivity. The RIHA test has been found to be a highly reproducible and sensitive method for detecting viral antigen in 5-10 ng of protein nitrogen. The sensitivity of the test was affected by the origins of the erythrocytes, i.e. from the different sheep from which they were drawn. To obtain results more rapidly, goose erythrocytes were used in place of sheep erythrocytes and the sensitized goose erythrocytes gave RIHA results in only 40 min.