Abstract: Methods and apparatus for speech recognition based on a hidden Markov model are disclosed. A disclosed method of speech recognition is based on a hidden Markov model in which words to be recognized are modeled as chains of states and trained using predefined speech data material. Known vocabulary is divided into first and second partial vocabularies where the first partial vocabulary is trained and transcribed using a whole word model and the second partial vocabulary is trained and transcribed using a phoneme-based model in order to obtain a mixed hidden Markov model. The transcriptions from the two models are stored in a single pronunciation lexicon and the mixed hidden Markov model stored in a singe search space. Apparatus are disclosed that also employ a hidden Markov model.

Abstract: A partial fragmentation bullet comprising a nose core, which forms a fragmenting portion of the partial fragmentation bullet, and a rear core, which forms a penetration portion of the partial fragmentation bullet. The nose core and the rear core are made of materials having the same level of hardness. This level of hardness is in the Brinell hardness range.

Abstract: A system and method for efficiently storing multiple versions of a data object in a database and generating a report to display a single version of a data object corresponding to any given time are provided. The data object is represented by a versioned data structure that includes a belief time field for storing a belief time corresponding to a version of the data object. A single version of the data object is retrieved by generating SQL code including a select clause for specifying the fields that are required for retrieving the data object and a where clause for specifying the version of the data object to be retrieved. A single foreign key is used to access multiple versions of the same data object.

Abstract: Systems and method for efficiently storing rate schedules and applying the rate schedules against operational events are provided. The rate schedules are represented in a database using a data structure that includes a field for storing a belief time during which the rate schedules are believed to be true. The database stores all versions of the rate schedules to ensure that complete audits of the rate schedules can be performed.

Abstract: A catalyst for oxidation of SO.sub.2 to SO.sub.3, which can be used as a fluidized bed catalyst. An abrasive resistant support is formed by adding silica filler and a clay binder to a silica sol; adding hydrated magnesium oxide to the resulting mixture, to form a gellable composition; dispersing the gellable composition in a liquid to form gel spheres; calcining the spheres; and treating them with an acid to remove acid soluble components. After the acid treatment, the support is impregnated with an alkali metal vanadate solution and calcined.