NetNews Usenet Archive 1992 #31

home *** CD-ROM | disk | FTP | other *** search

/ NetNews Usenet Archive 1992 #31 / NN_1992_31.iso / spool / comp / cogeng / 359 next >

Wrap

Text File | 1992-12-22 | 9.9 KB | 241 lines

Newsgroups: comp.cog-eng Path: sparky!uunet!tijc02!jdw255 From: jdw255@tijc02.uucp (John Wilson ) Subject: Speculation on the mind's data structures Message-ID: <1992Dec22.153529.29450@tijc02.uucp> Organization: Siemens Industrial Automation, Johnson City TN Distribution: net Date: Tue, 22 Dec 92 15:35:29 GMT Lines: 230 The Minds Eye This is speculation on the mechanics of data structures used by minds to think. From what I have read, memory and processing elements of the mind are interspersed and show confusing signs of locality of reference combined with global knowledge. Compactness and massive parallel processing are also remarked upon in the literature. Data seems to reside in general areas, not in any specific spot. I am speculating that these traits can be explained by an encoding scheme which creates an eye's viewpoint. What if the mind consists of a virtual space full of four dimensional objects. How would the information which defines this space be stored? How would processing elements look into this space? How would objects be manipulated? Four dimensional objects would consist of three spacial dimensions and one time dimension. How would the time dimension be handled? Four dimensions are hard to experiment with, so most of my work has been done in two dimensions with the expectation that it scales upwards. The experiments consisted of creating a one dimensional array from a two dimensional space which could be both stored and manipulated by parallel processing elements (PE's). In figure 1 sensor readings are given that a point of light would have at various distances. The value 6561 was chosen to make the numbers work out nicely. At a distance of 0 one sensor would be affected and it would receive all the light. At a distance of 8 seventeen sensors would be affected and would show a normal distribution of values adding up to 6561. Figures 2, 5, and 8 are diagrams which are converted to vectors in figures 3, 6, and 9 respectively. These vectors are converted to bar charts in figures 4, 7, and 10. The 50 element vectors hold information about a 450 point space. As the dimensions go up the compaction ratio also goes up. If the space is too crowded then information access becomes more difficult. Adding or removing points is accomplished by adding or subtracting appropriate values from figure 1 to a substring of the vector. Rotation of an object requires addition and subtraction of constants to bring some points closer while moving others away. Comparisions would be done by comparing substrings of vectors and might involve rotation or movement in the virtual space. Here are my speculative answers to my speculative questions! The information which defines the virtual space is stored as one dimensional vectors representing a viewpoint into the space. PE's look into this space by finding and processing subvectors. These subvectors contain the information about the points the PE's are interested in. Objects are manipulated by adding and subtracting values from appropriate subvectors. The time dimension would be handled by keeping old copies of the current vectors. Copies could be reduced in size to save space. This would cause some information to be lost. FIGURE 1 0 6561 1 2187 2187 2187 2 729 1458 2187 1458 729 3 243 729 1458 1701 1458 729 243 4 81 324 810 1296 1539 1296 810 324 81 5 27 135 405 810 1215 1377 1215 810 405 135 27 6 9 54 189 450 810 1134 1269 1134 810 450 189 54 9 7 3 21 84 231 483 798 1071 1179 1071 798 483 231 84 21 3 8 1 8 36 112 266 504 784 1016 1107 1016 784 504 266 112 36 8 1 FIGURE 2 9 * * * * * 8 7 * * * * 6 5 4 3 2 1 FIGURE 3 242 506 786 1052 1277 1452 1569 1604 1525 1347 1144 1016 1005 1074 1151 1206 1266 1368 1503 1611 1641 1584 1476 1361 1284 1275 1334 1417 1455 1417 1331 1257 1221 1214 1221 1254 1319 1380 1369 1261 1107 1004 999 1053 1067 972 768 517 291 123 FIGURE 4 . ... ... ... .... ..... . .... ..... ... . ..... ....... ..... ... ...... ................ ..... ...... ......................... ....... .......................... ........ ............................ .. .......................................... ........................................... ............................................. ............................................... ................................................ ................................................. .................................................. FIGURE 5 9 8 * 7 * 6 * 5 * 4 * 3 * 2 * 1 FIGURE 6 104 265 507 805 1100 1338 1499 1582 1584 1499 1372 1284 1301 1366 1380 1290 1218 1260 1404 1512 1548 1620 1701 1674 1566 1539 1782 1782 1458 1458 1701 2187 1458 729 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 FIGURE 7 . .. . .. .. .. .. ... .. .. .. ......... .. .... .............. .... . ............... ...... .. ............... ........... ................ ............................ ............................. .............................. ............................... ................................ ................................. .................................. FIGURE 8 9 8 7 * * 6 * * * * 5 * * * * * * 4 * * * * * * * 3 * * * * 2 * 1 FIGURE 9 1080 2646 4941 5589 6111 5940 6696 6822 6804 6129 5913 5790 6042 5889 5730 5505 5550 5661 5715 6024 6453 7209 7329 7002 5868 4728 3447 2367 1383 699 285 93 21 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 FIGURE 10 . .. ... .. ... ... ... ... .... . .... .... . .... . .... . .... . ..... ....... . ..... ....... .. ...... .......... ...... ........... ....... ........... ........ ............ ........ ............ ......... ...................... ....................... ........................ ......................... .......................... ........................... ............................ ............................. .............................. ............................... ................................ Best regards John Wilson