Abstract: An electronic device display may have pixels formed from crystalline semiconductor light-emitting diode dies, organic light-emitting diodes, or other pixel structures. The pixels may be formed on a display panel substrate. A display panel may extend continuously across the display or multiple display panels may be tiled in two dimensions to cover a larger display area. Interconnect substrates may have outwardly facing contacts that are electrically shorted to corresponding inwardly facing contacts such as inwardly facing metal pillars associated with the display panels. The interconnect substrates may be supported by glass layers. Integrated circuits may be embedded in the display panels and/or in the interconnect substrates. A display may have an active area with pixels that includes non-spline pixels in a non-spline display portion located above a straight edge of the display and spline pixel in a spline display portion located above a curved edge of the display.

Abstract: A morphological analysis method is described. In one aspect of the invention, the method includes the steps of breaking a word into potential base-affix pairs, modifying the base of each base-affix pair into possible roots, accessing each lexical entry of each root in a lexicon, deriving a lexical entry for a root if the lexical entry is not provided in the lexicon, retrieving the derivation information for each affix from the entry, applying the derivation information to the entry, and returning all derived entries.

Abstract: A sense discrimination system and method are described. One aspect of the present invention includes a method for performing sense discrimination, a token stream including a token wherein each sense of each lexical entry may be marked with a positive or negative morphology preference, each sense of each entry may be marked with relevant contexts, and each sense of each entry may be marked with a positive/negative status being provided, at least a portion of one parse tree and at least a portion of one head tail matrix also being provided. The method includes the steps of converting each preference marking into a raw numeric score, scoring the sense preference information, determining a best sense of a lexical entry from the sense preference information score, and inserting a designation for the best sense into a respective token.