"... In PAGE 1: ... These data were used to test the robustness of the map- ping algorithms to variations in the vocabulary and syntax used by the speakers to express the canonical phrases. Table1 shows two examples of the phrases and their alternatives. Each volunteer was also recorded speaking each of their alter- native phrases and these were subsequently transcribed by a speech recogniser and used to investigate the effects of recognition errors on the effectiveness of the mapping techniques.... In PAGE 3: ... A co-occurrence ma- trix is generated where each column represents a phrase and each row represent a word in a phrase. To illustrate this, the co-occurrence table for the example phrases in Table1 is shown in Table 3. As has been noted, simply using the counts of the words is not good for classification since the words occur with very different frequencies.... ..."

"... In PAGE 17: ...amount r1 1) (amount r2 2) ... (amount r6 3) Finally, action move is defined in a way that separates the resource requirements from the precondition and the effect lists: (:action move :parameters (?tr ?c1 ?c2 ?f) :precondition (and (truck ?tr) (city ?c1) (city ?c2) (at ?tr ?c1) (adjacent_cities ?c1 ?c2) (city_fuel ?c1 ?f)) :effect (and (not (at ?tr ?c1)) (at ?tr ?c2)) :resources (amount ?f 1)) Table2 shows the number of ground facts and ground actions for the first five problems of the mystery distribution, for the two alternative resource representations. As it is clear from this table, through the numerical representation of resources there is an important reduction in the number of ground facts, which is more considerable in the case of ground actions.... ..."

"... In PAGE 99: ... From #5B61#5D. are shown in Table9 . Standard #28additive binary#29 coding, in which each full split... In PAGE 100: ... Ronquist apos;s demonstration, that the phylogenetic information content of the matrices di#0Bers, does nothing to alter the redundancy of the information in Purvis apos;s sense. Indeed, Purvis apos;s modi#0Ccation must in some waychange the phylogenetic information con- tent of the matrix if it is to have the desired a#0Bect of reducing the apparent bias of standard MRP toward relationships in larger trees that he had identi#0Ced, or any a#0Bect at all! The point is illustrated using CIC as a measure of the information content of the elements #28each of which corresponds to a tree with a single inter- nal branch#29 in the alternative matrix representations in Table9 . The information content of the larger source tree, tree 41#28a#29, is less than the sum of the CICs of the matrix elements in standard MRP #283:91 bits versus 2:32 + 2:32 = 4:64 bits, respectively#29 indicating some redundancy.... In PAGE 101: ...Supertrees 90 information it does not, as he claimed, remove all the redundant information. Con- sider the second element under Purvis coding, #28 Table9 , element 4#29. This partial split completely resolves the relationships of leaves A, B and C.... In PAGE 101: ... That D is basal to A, B and C is conveyed by the other element, element3. However, this elementwould convey this information if either A or B were coded as missing entries, producing two al- ternative minimal matrices #28 Table9 , elements 4 and 5 and elements 4 and 6#29. The minimal matrices yield the original source tree when analyzed with parsimonyso the information removed is redundant in Purvis apos;s sense.... ..."

"... In PAGE 9: ... fth alternative, associating valid time with sets of tuples (i.e., relations) or object graphs, has not been incorporated into any of the proposed data models, primarily because it lends itself to high data redundancy. Table3 categorizes most the the data models along these two dimensions. We do not include the OODAPLEX, Sciore-1, and TIGUKAT data models as these two aspects are arbitrarily speci able in these models.... ..."

"... In PAGE 18: ... Another issue concerns whether transaction time is associated with individual attribute values, with tuples, or with sets of tuples. The choices made in the various data models are characterized in Table3 . Gadia-3 is the only data model to timestamp attribute values; it is di cult to e ciently implement this alternative directly.... In PAGE 26: ... In Table 2, the BCDM occupies the un lled entry corresponding to timestamping tuples with valid-time elements. In Table3 , the BCDM occupies the un lled entry corresponding to times- tamping tuples with transaction-time elements. Hence, the BCDM is unique in that it timestamps tuples with bitemporal elements.... ..."

"... In PAGE 17: ...ribute values? A third alternative, associating valid time with sets of tuples, i.e., relations, has not been incorporated into any of the proposed data models, primarily because it lends itself to high data redundancy. The data models are evaluated on these two aspects in Table2 . Interestingly, only one quadrant, timestamping tuples with an valid-time element, has not been considered.... In PAGE 26: ... The timestamps are temporal elements, as in Cli ord-2, Gadia-1 and Gadia- 3. In Table2 , the BCDM occupies the un lled entry corresponding to timestamping tuples with valid-time elements. In Table 3, the BCDM occupies the un lled entry corresponding to times- tamping tuples with transaction-time elements.... ..."