File size: 116,005 Bytes
6fa4bc9 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 | {
"paper_id": "2021",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T13:27:06.417825Z"
},
"title": "Discourse-Driven Integrated Dialogue Development Environment for Open-Domain Dialogue Systems",
"authors": [
{
"first": "Denis",
"middle": [],
"last": "Kuznetsov",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Moscow \u2021 AIR Institute",
"location": {
"settlement": "Moscow"
}
},
"email": "kuznetsov.dp@phystech.edu"
},
{
"first": "Dmitry",
"middle": [],
"last": "Evseev",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Moscow \u2021 AIR Institute",
"location": {
"settlement": "Moscow"
}
},
"email": "dmitrii.evseev@phystech.edu"
},
{
"first": "Lidiya",
"middle": [],
"last": "Ostyakova",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Moscow \u2021 AIR Institute",
"location": {
"settlement": "Moscow"
}
},
"email": "l.ostyakova@yandex.ru"
},
{
"first": "Oleg",
"middle": [],
"last": "Serikov",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Moscow \u2021 AIR Institute",
"location": {
"settlement": "Moscow"
}
},
"email": "oserikov@hse.ru"
},
{
"first": "Daniel",
"middle": [],
"last": "Kornev",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Moscow \u2021 AIR Institute",
"location": {
"settlement": "Moscow"
}
},
"email": "danielko@deeppavlov.ai"
},
{
"first": "Mikhail",
"middle": [],
"last": "Burtsev",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Moscow \u2021 AIR Institute",
"location": {
"settlement": "Moscow"
}
},
"email": "burtcev.ms@mipt.ru"
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "Development environments for spoken dialogue systems are popular today because they enable rapid creation of the dialogue systems in times when usage of the voice AI Assistants is constantly growing. We describe a graphical Discourse-Driven Integrated Dialogue Development Environment (DD-IDDE) for spoken open-domain dialogue systems. The DD-IDDE allows dialogue architects to interactively define dialogue flows of their skills/chatbots with the aid of the discourse-driven recommendation system, enhance these flows in the Python-based DSL, deploy, and then further improve based on the skills/chatbots usage statistics. We show how these skills/chatbots can be specified through a graphical user interface within the VS Code Extension, and then run on top of the Dialog Flow Framework (DFF). An earlier version of this framework has been adopted in one of the Alexa Prize 4 socialbots while the updated version was specifically designed to power the described DD-IDDE solution.",
"pdf_parse": {
"paper_id": "2021",
"_pdf_hash": "",
"abstract": [
{
"text": "Development environments for spoken dialogue systems are popular today because they enable rapid creation of the dialogue systems in times when usage of the voice AI Assistants is constantly growing. We describe a graphical Discourse-Driven Integrated Dialogue Development Environment (DD-IDDE) for spoken open-domain dialogue systems. The DD-IDDE allows dialogue architects to interactively define dialogue flows of their skills/chatbots with the aid of the discourse-driven recommendation system, enhance these flows in the Python-based DSL, deploy, and then further improve based on the skills/chatbots usage statistics. We show how these skills/chatbots can be specified through a graphical user interface within the VS Code Extension, and then run on top of the Dialog Flow Framework (DFF). An earlier version of this framework has been adopted in one of the Alexa Prize 4 socialbots while the updated version was specifically designed to power the described DD-IDDE solution.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "Conversational AI focused on the development of the dialogue systems is one of the most challenging areas in the artificial intelligence given the subjectivity of the human language interpretation, and is gaining more and more popularity across both academia (Viewer, 2021) and consumer & enterprise markets (Statista, 2020) .",
"cite_spans": [
{
"start": 308,
"end": 324,
"text": "(Statista, 2020)",
"ref_id": "BIBREF27"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Dialogue systems are classified into taskoriented dialogue systems and chat (or non-taskoriented) dialogue systems. Usually they are studied differently, but combining them has been proposed (Lee et al., 2006) , and has been found effective in improving user impressions and the relationships with users (Lucas et al., 2018) . While taskoriented systems are more widespread and used in different business scenarios (Bocklisch et al., 2017) , building systems that can carry multi-turn open-domain conversations is still far from a solved problem (Hu et al., 2021) .",
"cite_spans": [
{
"start": 191,
"end": 209,
"text": "(Lee et al., 2006)",
"ref_id": "BIBREF19"
},
{
"start": 304,
"end": 324,
"text": "(Lucas et al., 2018)",
"ref_id": "BIBREF21"
},
{
"start": 415,
"end": 439,
"text": "(Bocklisch et al., 2017)",
"ref_id": "BIBREF3"
},
{
"start": 546,
"end": 563,
"text": "(Hu et al., 2021)",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "While there is a significant effort in using large generative models (Brown et al., 2020) , (Roller et al., 2020) in the open-domain dialogue systems, recent tests of GPT-3 (Marcus and Davis, 2020) , and Conversational AI challenges like Alexa Prize Socialbot Grand Challenge (Ram et al., 2018) highlight the inherent weaknesses of these models, namely inability to maintain context, avoid contradictions, and control the dialogue strategically in the long dialogues. These fallacies experienced by us and other teams during the Amazon Alexa Prize 3 (Gabriel et al., 2020) led us to the importance of continuing research towards the tightly-controlled open-domain dialogue management frameworks like E-STDM (Finch and Choi, 2020) .",
"cite_spans": [
{
"start": 69,
"end": 89,
"text": "(Brown et al., 2020)",
"ref_id": null
},
{
"start": 92,
"end": 113,
"text": "(Roller et al., 2020)",
"ref_id": null
},
{
"start": 167,
"end": 197,
"text": "GPT-3 (Marcus and Davis, 2020)",
"ref_id": null
},
{
"start": 276,
"end": 294,
"text": "(Ram et al., 2018)",
"ref_id": "BIBREF24"
},
{
"start": 550,
"end": 572,
"text": "(Gabriel et al., 2020)",
"ref_id": null
},
{
"start": 707,
"end": 729,
"text": "(Finch and Choi, 2020)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "During the Amazon Alexa Prize 4 (Hu et al., 2021) competition we developed (Baymurzina et al., 2021) a first iteration of such framework called Dialog Flow Framework (DFF) (Kuznetsov and Evseev, 2021) to aid in rapid development of open-domain \"skills\" covering various societal topics in our DREAM Socialbot. Usage showed that while users enjoy controlled conversation in their favorite domains, they often interrupt the conversation with the out-of-domain (Konr\u00e1d et al., 2021) responses leading it into the non-anticipated directions.",
"cite_spans": [
{
"start": 75,
"end": 100,
"text": "(Baymurzina et al., 2021)",
"ref_id": null
},
{
"start": 172,
"end": 200,
"text": "(Kuznetsov and Evseev, 2021)",
"ref_id": "BIBREF17"
},
{
"start": 458,
"end": 479,
"text": "(Konr\u00e1d et al., 2021)",
"ref_id": "BIBREF15"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Feedback of our Alexa Prize 4 team members who wrote DFF-based open-domain \"skills\" shows that predictions of the possible user response could significantly cut the time needed by the developer to design and develop these skills.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "The key insight into the conversation is that each utterance in the dialogue is also a kind of action being performed by the speaker (Austin, 1962) . These actions are called speech or dialog acts (Jurafsky and Martin, 2009 ). Yet while dialog acts used by some of the Alexa Prize teams e.g., MIDAS (Yu et al., 2019) , are quite sophisticated in understand-ing higher level user intents conceptually similar to those used in task-oriented systems (Bocklisch et al., 2017) , they are single turn-based and don't support the development of the open-domain conversations that are interactional and sequential (Eggins and Slade, 1996) . Indeed, these dialog act taxonomies lack connection to the discourse level of the casual conversation analysis as shown by (Halliday and M., 2004) .",
"cite_spans": [
{
"start": 133,
"end": 147,
"text": "(Austin, 1962)",
"ref_id": "BIBREF1"
},
{
"start": 197,
"end": 223,
"text": "(Jurafsky and Martin, 2009",
"ref_id": "BIBREF14"
},
{
"start": 299,
"end": 316,
"text": "(Yu et al., 2019)",
"ref_id": "BIBREF32"
},
{
"start": 447,
"end": 471,
"text": "(Bocklisch et al., 2017)",
"ref_id": "BIBREF3"
},
{
"start": 606,
"end": 630,
"text": "(Eggins and Slade, 1996)",
"ref_id": "BIBREF9"
},
{
"start": 756,
"end": 779,
"text": "(Halliday and M., 2004)",
"ref_id": "BIBREF12"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In the effort of making it easier for developers to design and develop sequential and interactional multi-turn open-domain dialogue \"skills\" we propose a new graphical Discourse-Driven Integrated Dialogue Development Environment (DD-IDDE) that combines a novel Discourse-Driven Recommendation System and a DSL for the Dialog Flow Framework, encapsulated together into an extension to the popular IDE Visual Studio Code.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "One of the first Development Environments designed for spoken language systems, (Denecke, 2000) , was created in 2000. While it supported an entire dialogue system development lifecycle, it lacked the power of the modern NLP components such as NER, Entity Linking, etc. as well as the means for the development of the open-dialogue skills/chatbots. Today, there is a variety of dialogue development frameworks that have emerged to aid in rapid dialogue system creation. Rule-based approaches used e.g. in (Bocklisch et al., 2017) are common in the task-oriented systems broadly used in businesses. Other systems incorporate support for full development lifecycle of the dialogue systems starting with individual NLP ML models and ending with the complete chatbots (Burtsev et al., 2018) . Some of these systems are also encapsulated in the form of web-based IDEs e.g. (Lawrence et al., 2021) .",
"cite_spans": [
{
"start": 80,
"end": 95,
"text": "(Denecke, 2000)",
"ref_id": "BIBREF7"
},
{
"start": 505,
"end": 529,
"text": "(Bocklisch et al., 2017)",
"ref_id": "BIBREF3"
},
{
"start": 764,
"end": 786,
"text": "(Burtsev et al., 2018)",
"ref_id": "BIBREF6"
},
{
"start": 868,
"end": 891,
"text": "(Lawrence et al., 2021)",
"ref_id": "BIBREF18"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "Domain-specific languages (DSLs) are typically used in dialogue systems to aid in rapid layout of the dialogue logic. One of the oldest DSLs in this space, AIML (Wallace, 2001 ) was created as an XML dialect that supports pattern matching of the user utterance and template-based entity extraction and was used in development of the famous opendomain dialogue system \"A.L.I.C.E.\" (Wallace, 2009) . RASA DSL (Bocklisch et al., 2017) was designed for the development of the task-oriented systems and includes NLU (intent detection and entity extraction) and stories (actions for intents) components. JAICF (AI, 2021), a JavaScript-based state machine-based DSL, was designed to support both task-oriented systems and chatbots.",
"cite_spans": [
{
"start": 161,
"end": 175,
"text": "(Wallace, 2001",
"ref_id": "BIBREF30"
},
{
"start": 380,
"end": 395,
"text": "(Wallace, 2009)",
"ref_id": "BIBREF31"
},
{
"start": 407,
"end": 431,
"text": "(Bocklisch et al., 2017)",
"ref_id": "BIBREF3"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "A winner of Alexa Prize 2, Gunrock team enhanced their socialbot's (Yu et al., 2019) internal Dialog Flow framework with the MIDAS dialog act scheme to create an adaptive and unique conversation experience with the user. However it does not enable bot developers with the mechanisms to manage dialogue strategically across multiple conversation turns. Slugbot (Bowden et al., 2019) , another Alexa Prize team, proposed a DRDM dialogue model to control the coherence of the open-domain dialogue using discourse relations. Their approach introduced a combination of dialog acts and four discourse relations from the PDTB 2.0 (Prasad et al., 2008 ) as means to model interaction within individual turns and at a higher level. Nonetheless, PDTB 2.0 is based on the 1-million-word Wall Street Journal corpus which is a written language and is not best suited for the spoken casual conversation analysis.",
"cite_spans": [
{
"start": 67,
"end": 84,
"text": "(Yu et al., 2019)",
"ref_id": "BIBREF32"
},
{
"start": 360,
"end": 381,
"text": "(Bowden et al., 2019)",
"ref_id": "BIBREF4"
},
{
"start": 623,
"end": 643,
"text": "(Prasad et al., 2008",
"ref_id": "BIBREF23"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "DD-IDDE is powered by the Dialog Flow Framework (DFF) derived from the E-STDM framework which is conceptually similar to both Dialog Flow framework described in (Yu et al., 2019) and the DRDM-dialogue model. It's newly developed DSL was specifically created so that: (1) all conditions, responses and even dialogue flows could be implemented as Python functions, and (2) conditions for state transition could use any of the available annotators like Speech Function Classifier and Predictor and others within the larger DREAM socialbot or derived chatbots and assistants. However, while it is relatively common for dialogue flow frameworks to have a built-in DSL and support pluggable NLU components, DD-IDDE specifically incorporates the support for the novel Discourse-Driven Recommendation System as well as the visual tools aiding in designing open-domain dialogues across multiple conversation steps.",
"cite_spans": [
{
"start": 161,
"end": 178,
"text": "(Yu et al., 2019)",
"ref_id": "BIBREF32"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "Development of the discourse-driven open-domain dialogue \"skills\" and chatbots is an ongoing iterative cycle that includes: (1) design of the \"skill\"/chatbot as the sequence of conditional transitions between states, aided by the discourse moves recommendations, (2) development of the \"skill\"/chatbot, (3) deployment, ending with the (4) analysis of the running solution, followed by the restart of the cycle.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Concepts",
"sec_num": "3"
},
{
"text": "Flows, Nodes, Transitions, Conditions and Responses: We split the dialogue graph into the logically related parts called Flows to make it more convenient to work with. Each Flow represents a dialogue subgraph. Typically, flows can be organized to cover different topics in the dialogue graph. Like in (Finch and Choi, 2020) , we identify the states of the dialogue at a certain moment, such states correspond to the Nodes of the dialogue graph. Being in a certain node of the graph, we can go to an another node if the Condition that corresponds to this Transition is true. After each transition, the Responses corresponding to the node are returned.",
"cite_spans": [
{
"start": 301,
"end": 323,
"text": "(Finch and Choi, 2020)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Concepts",
"sec_num": "3"
},
{
"text": "While other solutions often incorporate dialogue/speech acts (DAMSL-SWBD, ISO standard, HCRC etc.) to recognize the higher-level user intentions in a particular context and describe discourse of conversations, they still lack the ability to represent conversational structure and make it impossible to analyze discourse at the dialogue level. In DD-IDDE, we propose implementation of more enhanced intents called Speech Functions that not only represent single turns but also reflect their role in the larger dialogue context.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Functions and Discourse Moves:",
"sec_num": null
},
{
"text": "Eggins and Slade in (Eggins and Slade, 1996) introduced an approach connecting dialogue turns and cross-turn discourse structure patterns specific for casual conversation as the higher-level abstraction. At turn level, they extended Halliday's concept of Speech Functions (Halliday and M., 2004) that express pragmatic goals of speakers and can be used as an enhanced alternative to Dialogue and Speech Acts. Advantages of Speech Functions are in specified grammatical criteria for Speech Function identification and the ability to constrain a comprehensive systematic discourse model of dialogues with the help of them.",
"cite_spans": [
{
"start": 20,
"end": 44,
"text": "(Eggins and Slade, 1996)",
"ref_id": "BIBREF9"
},
{
"start": 272,
"end": 295,
"text": "(Halliday and M., 2004)",
"ref_id": "BIBREF12"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Functions and Discourse Moves:",
"sec_num": null
},
{
"text": "These Speech Functions express different discourse purposes and can compound patterns that are expressed by so-called moves. There are three types of Discourse Moves as defined by Eggins and Slade: (i) Opening moves are used to start a discussion of a new topic in the dialogue. (ii) Sustaining moves develop the current topic of conversation. (iii) Reacting moves denote the transfer of the role of the current speaker. The first subtype of reacting moves are responses (React.Response), which lead dialogues or top-ics of conversation to completion, the second one is rejoinders (React.Rejoinder), which, on the contrary, prolong the conversation.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Functions and Discourse Moves:",
"sec_num": null
},
{
"text": "To enable users design dialogue flows with the aid of the Speech Functions and Discourse Moves, we built Speech Function Classifier (4.2) and Speech Function Predictor (4.3).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Functions and Discourse Moves:",
"sec_num": null
},
{
"text": "DFF DSL: DFF DSL is Python-based language for definition of a dialogue graph, conditions of transitions between nodes (e.g. checks for Speech Functions), and functions for processing user or dialogue system utterances (e.g. entity extraction). All the elements in DFF DSL can be callable (functions), including dialogue graphs, which makes the DSL flexible and extensible.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Functions and Discourse Moves:",
"sec_num": null
},
{
"text": "DFF Extensions: DFF DSL extensions are modules which can be used in the dialog flow for NLU or NLG. NLU extensions include Speech Function Classifier, Speech Function Predictor and Entity Extraction Module. DFF DSL extensions for NLG are Slot Filling Module (fills the dialogue system's response template with extracted entities), Factoid Response Generation Module (provides content of pages from text databases) and Generic Responses Module (returns short utterances such as \"Yes\" or \"No\", which represent appropriate Speech Function responses to user utterance's Speech Function).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Functions and Discourse Moves:",
"sec_num": null
},
{
"text": "We use Dialog Flow Framework (DFF) as a core of the DD-IDDE. Creating conversational agents with the support of the open-domain skills is challenging as was shown, for example, in (Kuratov et al., 2019) . Currently, there are many ready-made solutions (Finch and Choi, 2020) , (Lison and Kennington, 2016) , (Ultes et al., 2017) , (Bocklisch et al., 2017) .",
"cite_spans": [
{
"start": 180,
"end": 202,
"text": "(Kuratov et al., 2019)",
"ref_id": "BIBREF16"
},
{
"start": 252,
"end": 274,
"text": "(Finch and Choi, 2020)",
"ref_id": "BIBREF10"
},
{
"start": 277,
"end": 305,
"text": "(Lison and Kennington, 2016)",
"ref_id": "BIBREF20"
},
{
"start": 308,
"end": 328,
"text": "(Ultes et al., 2017)",
"ref_id": "BIBREF28"
},
{
"start": 331,
"end": 355,
"text": "(Bocklisch et al., 2017)",
"ref_id": "BIBREF3"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Core",
"sec_num": "4.1"
},
{
"text": "The alpha version of DFF was developed during the Alexa Prize 4 Challenge, based on E-STDM (Finch and Choi, 2020) . After the Challenge, all the shortcomings formulated during the competition were taken into account and the DFF DSL was rewritten. DFF (Kuznetsov and Evseev, 2021) is flexible, expressive, minimal, easy to expand through ready-made extensions or own extensions, as well as collect statistics with subsequent analysis.",
"cite_spans": [
{
"start": 91,
"end": 113,
"text": "(Finch and Choi, 2020)",
"ref_id": "BIBREF10"
},
{
"start": 251,
"end": 279,
"text": "(Kuznetsov and Evseev, 2021)",
"ref_id": "BIBREF17"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Core",
"sec_num": "4.1"
},
{
"text": "Having minimal dependencies, DFF is lightweight and stateless. This allows it to be used directly as a service in the case of the microservice architecture applications or as a separate API accessible from outside.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Core",
"sec_num": "4.1"
},
{
"text": "DFF has an expressive Python-based DSL called DFF DSL specially designed to be a minimalistic and extensible scripting markup language when compared to other solutions (e.g., AIML, RiveScript, ChatScript, botml), which supports using specific extensions (e.g., discourse-driven integrated dialogue development, a mechanism to extract and use entities based on their Wikidata, generic responses based on the speech functions ontology and etc).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Core",
"sec_num": "4.1"
},
{
"text": "There is a set of community-created extentions that are well documented, and it is straightforward to add own extentions to the framework.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Core",
"sec_num": "4.1"
},
{
"text": "To aid in the development of the Speech Function Classifier, we picked the Santa Barbara Corpus of Spoken American English, which consists of 60 transcriptions of the naturally-occurring spoken conversations. Three face-to-face dialogues were preprocessed and then labeled with the Speech Functions into a small dataset including about 1700 manually annotated utterances. Two annotators reached an inter-annotator agreement of \u03ba = 0.71 on 1200 utterances which is considered to be a good result. By limiting taxonomy from 45 to 33 classes, using a hierarchical algorithm based on several Logistic Regression models with different parameters, and a rule-based approach, the last version of Speech Function Classifier achieved an F1-score varying from 52% to 71% depending on the distribution of the Speech Functions in a particular dialogue. The resulting Speech Function Classifier annotator classifies each phrase in the user's utterance as well as each dialogue system's response candidate with Speech Functions. This classification enables the dialogue system to predict the dialogue system's response Speech Functions most expected by the user.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Function Classifier",
"sec_num": "4.2"
},
{
"text": "Speech Function Predictor yields probabilities of speech functions that can follow a speech function predicted by Speech Function Classifier. Speech Function Predictor is a model based on the statistics of speech function sequences in the manually annotated data. As the collected data is not fully representative, rules were added covering missing cases of speech functions sequences. Speech Func-tion Predictor annotation to each last phrase of the user's utterances serves as a recommendation for a dialogue system's next response. It is deployed as the HTTP endpoint available for the VS Code extension as the Recommendation API.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Speech Function Predictor",
"sec_num": "4.3"
},
{
"text": "The recommendation system uses Speech Functions used to classify each bot utterance as an input for the Speech Function Predictor to suggest the possible kinds of user responses. The same mechanism is then used to design next bot responses based on the user responses, enabling the rapid design of the sequential discourse-driven opendomain dialogue system.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discourse-Driven Recommendation System",
"sec_num": "4.4"
},
{
"text": "The DD-IDDE offers a split GUI (Serikov and Babadeev, 2021) that offers both visual design of the current open-domain skill's/chatbot's dialogue flows and a mechanism for working with the codebehind that represents the logic of the aforementioned skill/chatbot. It also provides a mechanism to obtain recommendations for the discourse moves for the selected node(s). Visual dialogue graph editor allows users to sketch DFF scenarios in form of the dialogue graph. Here, the nodes of the graph are System States, the transitions are allowed by users' utterances (see figure 4) . It is powered by the modified Draw.IO (JGraph Ltd, 2021), (Dieterichs and Rouill\u00e9, 2021) editor interface.",
"cite_spans": [
{
"start": 636,
"end": 666,
"text": "(Dieterichs and Rouill\u00e9, 2021)",
"ref_id": "BIBREF8"
}
],
"ref_spans": [
{
"start": 566,
"end": 575,
"text": "figure 4)",
"ref_id": "FIGREF2"
}
],
"eq_spans": [],
"section": "Graphical User Interface",
"sec_num": "4.5"
},
{
"text": "Code behind editor is a Python editor, and it shows DFF DSL produced from the dialogue flows designed in the visual editor. It can be extended with the standard Pylint-based recommendations and aids user in fixing errors prior to launching the target skill/chatbot. This GUI is provided in the form of the VS Code Extension.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Graphical User Interface",
"sec_num": "4.5"
},
{
"text": "The DD-IDDE enables users to further improve their skills/chatbots based on the analysis of the user behavior. A DFF Extension was made that allows collecting the particular dialogue flows and individual nodes usage statistics at run-time as well as showing statistics charts in a dashboard.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conversational Analysis",
"sec_num": "4.6"
},
{
"text": "The aforementioned DFF Extension for Conversation Analysis enables DFF open-domain \"skill\" developers to use the collected statistics of frequencies of transitions between nodes to focus developer's attention on the user behavior. Developing With and Without Discourse-Driven Recommendations: Development of DFF DSL-based open-domain \"skill\" is a multi-step cycle that includes initial skill design, development, test, deployment, feedback gathering, until skill quality is improved. The most challenging parts of the skill design are: (1) prediction of the out-of-domain responses that can be not so obvious for developers, and (2) sequential dialogue design. Our initial experiments with the DD-IDDE and its Discourse Moves recommendations showed that time required to address both the out-of-domain responses and to design sequential dialogue was significantly cut compared to the pure DFF DSL skills design.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "General Statistical Recommendation System",
"sec_num": "4.7"
},
{
"text": "In this paper, we introduced a new Discourse-Driven Integrated Dialogue Development Environment (DD-IDDE) focused on aiding users in building open-domain skills/chatbots based on the discourse-driven recommendations. Our DD-IDDE makes it possible for novice users to rapidly design their scenario-driven skills in the visual editor with the discourse move recommendations, and then expand their skills with the custom Python code. In the future versions we plan to further expand the DD-IDDE to enable seamless transitions between the manually edited DFF DSL and its visual graph representation, add deeper integration of the statistics into the main interface of the VS Code extension, as well as add other extentions to the underlying DFF framework.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusions and Future Directions",
"sec_num": null
},
{
"text": "Having installed the DD-IDDE VS Code Extension 1 , user creates a new .drawio file in the common/dff_markup_scenarios with the name of the target skill scenario, e.g., pet_scenario.drawio -see 1.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.1 Designing",
"sec_num": null
},
{
"text": "User then proceeds with defining the Speech Function of the starting node by double-clicking on it -see 2.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.1 Designing",
"sec_num": null
},
{
"text": "Once the first node has been set up, user can use the Speech Function Recommendations to pick the next discourse moves by clicking on the \"Show Suggestions\" menu item -see 3.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.1 Designing",
"sec_num": null
},
{
"text": "User then continues adding nodes based on their scenario, with the aid of the Speech Function recommendations -see 4.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.1 Designing",
"sec_num": null
},
{
"text": "Once the designed skill is sufficiently outlined, user proceeds with the use of the Show Markup functionality that generates the code based on the designed skill -see 5.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.2 Developing",
"sec_num": null
},
{
"text": "With the generated DFF DSL code, user can now add the additional code-behind functionality such as the use of the regular expressions, custom intents, and the like -see 6.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.2 Developing",
"sec_num": null
},
{
"text": "In the Conversational Analysis dashboard user can observe the frequency of nodes' visits of the individual dialogue flows, rates of transitions within dialogues, as well as the frequency of the user responses. User can also collect statistics of annotations of user and bot utterances, custom data coming from other DFF extentions, e.g., classification of user utterances with the speech functions, user and bot emotions etc. The dashboard is described in more detail in appendix D. The dashboard allows users to select as many dialogues as one dialog and show user behavior on it. On this dialog graph is shown for one dialog 12 and on this one for many 8.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.3 Analyzing",
"sec_num": null
},
{
"text": "Interface Illustrations VS Code Extension GUI is shown in figures 1, 2, 3, 4, 5, 6 DFF as well as many extentions for it are distributed with an open source license, like many other frameworks (e.g., Emora STDM, AIML, RiveScript, ChatScript, botml, OpenDial, PyDial). DFF is distributed with Apache 2.0 license. DFF DSL is a Python-based DSL for designing scenario-driven skills/chatbots. The main element of a skill in DFF DSL format is a dialogflow (a set of nodes and transitions between them, which refers to a particular subtopic). A dialogflow is a Python dictionary where keys are node names and elements describe response in this node and transitions to other nodes in the same dialogflow or in other dialogflows (Fig. 13) . The main features of a dialogflow in DFF DSL format:",
"cite_spans": [],
"ref_spans": [
{
"start": 721,
"end": 730,
"text": "(Fig. 13)",
"ref_id": "FIGREF0"
}
],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "\u2022 a condition for transition between nodes can be one of predefined conditions in DFF DSL (for example, a regular expression or a function which checks the speech function in the user utterance, etc.)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "\u2022 a response can be a string or a function",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "\u2022 processing of user utterance and response with custom functions (which can be used, for example, for slot extraction and slot filling (Fig. 14) )",
"cite_spans": [],
"ref_spans": [
{
"start": 136,
"end": 145,
"text": "(Fig. 14)",
"ref_id": "FIGREF0"
}
],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "\u2022 extensions for using facts databases (for example, Wikipedia) for response generation (Fig. 16 ) or generic responses to user utterance speech function",
"cite_spans": [],
"ref_spans": [
{
"start": 88,
"end": 96,
"text": "(Fig. 16",
"ref_id": "FIGREF0"
}
],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "The solution, like many others (e.g., AIML), is based on the state machine. However, unlike others, DFF aids in the development of the open-domain skills/chatbots with the help of the discourse moves recommendation making it easy to design these skills and chatbots. It also incorporates a mechanism to extract and use entities based on their Wikidata types, as well as to provide generic responses based on the speech functions ontology.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "DFF was designed to be extensible as a modular system. With the external extentions, DFF can be significantly expanded as in complex frameworks (Finch and Choi, 2020) , (Bocklisch et al., 2017) . There is a set of community-created extentions that are well documented, and it is straightforward to add own extentions to the framework. DFF DSL is automatically validated before starting the bot, making it easy to identify issues beforehand..",
"cite_spans": [
{
"start": 144,
"end": 166,
"text": "(Finch and Choi, 2020)",
"ref_id": "BIBREF10"
},
{
"start": 169,
"end": 193,
"text": "(Bocklisch et al., 2017)",
"ref_id": "BIBREF3"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "A.4 VS Code Extension Graphical User",
"sec_num": null
},
{
"text": "The target node for transition can be:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "B.2 Transitions",
"sec_num": null
},
{
"text": "\u2022 the node name from the current dialogflow;",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "B.2 Transitions",
"sec_num": null
},
{
"text": "\u2022 the name of another dialogflow and the node in this dialogflow;",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "B.2 Transitions",
"sec_num": null
},
{
"text": "\u2022 special keywords (\"forward\" -to the next node in the current dialogflow, \"back\" -to the previous node, \"repeat\" -to make cycle transition, \"previous\" -to the previous active node).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "B.2 Transitions",
"sec_num": null
},
{
"text": "Transitions between nodes can be local (which can be used only in the dialogflow where they were defined) or global (can be used in any other dialogflow).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "B.2 Transitions",
"sec_num": null
},
{
"text": "The approach of dividing DFF into modules allows users to incorporate both lightweight annotators, classifiers, and so on, as well as modules based on neural models that require a large amount of resources in their solutions.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C DFF: Extensions",
"sec_num": null
},
{
"text": "In production ready systems, it is necessary to use keywords and a matching pattern for the agent to work, as well as the selection of entities, the presence of ontologies, the identification of intents, the classification of sentiments, and much more.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C DFF: Extensions",
"sec_num": null
},
{
"text": "Generic responses are short utterances, generated by the bot in response to a particular speech function in user utterance. Generic responses can be useful if the user is proactive and responds with long and detailed utterances, the bot in this case will say somewhat like \"yes\", \"I agree\". Examples of generic responses:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.1 Generic Responses Module",
"sec_num": null
},
{
"text": "\u2022 \"No\" in response to \"React.Respond.Confront.Reply.Contradict\" and \"React.Respond.Reply.Contradict\" speech functions;",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.1 Generic Responses Module",
"sec_num": null
},
{
"text": "\u2022 \"I don't know.\" to \"React.Respond.Confront.Reply.Disawow\" and \"React.Respond.Reply.Disawow\";",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.1 Generic Responses Module",
"sec_num": null
},
{
"text": "\u2022 \"I don't agree with you\" to \"React.Respond.Confront.Reply.Disagree\" and \"React.Respond.Reply.Disagree\";",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.1 Generic Responses Module",
"sec_num": null
},
{
"text": "\u2022 \"Yes\" to \"React.Respond.Support.Reply.Affirm\", \"React.Respond.Support.Reply.Agree\" and \"React.Respond.Reply.Agree\".",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.1 Generic Responses Module",
"sec_num": null
},
{
"text": "Generic responses are implemented as an extention for DFF DSL with one cycled node of dialogflow and a response function which extracts a speech function from the user utterance and generates a suitable generic response.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.1 Generic Responses Module",
"sec_num": null
},
{
"text": "Entity Extraction Module is used to find entities of particular types in the user utterance and store them in shared memory. Entities can be extracted by:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.2 Other Extentions",
"sec_num": null
},
{
"text": "\u2022 Wikidata Entity Type: user can specify a Wikidata ID for the given entity type to find entities of this type in the given utterance, e.g. to find a city user should specify \"wiki:Q515\" (which corresponds to \"city\" in Wikipedia)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.2 Other Extentions",
"sec_num": null
},
{
"text": "\u2022 Entity Tag from the set of (\"PERSON\", \"LO-CATION\", \"ORGANIZATION\", etc.): user can specify one of these tags to find entities of this type, e.g., to find a location user should specify \"tag:LOCATION\" Slot Filling Module is used to fill extracted entities in the slots of the response.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.2 Other Extentions",
"sec_num": null
},
{
"text": "Factoid Response Generation Module provides content of the page from the databases like Wikipedia, wikiHow, etc. mentioned as the argument of the extension function.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.2 Other Extentions",
"sec_num": null
},
{
"text": "Generic Responses Module returns short utterances (for example, \"Yes\", \"No\", \"I agree\" etc.) that represent selected speech functions as appropriate response to the selected user utterance's speech functions.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C.2 Other Extentions",
"sec_num": null
},
{
"text": "The statistics DFF extention 2 comes with a dashboard that displays various statistics. Using them, users can evaluate the user's use of script branches. The dashboard is shown in figures 7, 8, 9, 10, 12 The abscissa axis shows the order of a step in the dialog, the ordinate axis shows the node in which the user is located for this step. ",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "D DFF: Conversational Analysis Dashboard",
"sec_num": null
},
{
"text": "https://github.com/deepmipt/vscode-dff/ releases/tag/0.2.1",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "https://github.com/kudep/dff-node-stats/",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
}
],
"back_matter": [
{
"text": "Authors are deeply grateful to the Alexa Prize organizers for their feedback and advice during the Alexa Prize Grand Challenge 4. Authors also thanks our contributor Dmitry Babadeev who became an owner of the VS Code extension encapsulating the DD-IDDE, and all members of Neural Networks and Deep Learning Lab for their support and advice during the project development. Oleg Serikov is partially supported by the framework of the HSE University Basic Research Program.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Acknowledgements",
"sec_num": null
}
],
"bib_entries": {
"BIBREF1": {
"ref_id": "b1",
"title": "How to do things with words",
"authors": [
{
"first": "Austin",
"middle": [],
"last": "John Langshaw",
"suffix": ""
}
],
"year": 1962,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "John Langshaw Austin. 1962. How to do things with words. William James Lectures. Oxford University Press.",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Rasa: Open source language understanding and dialogue management",
"authors": [
{
"first": "Tom",
"middle": [],
"last": "Bocklisch",
"suffix": ""
},
{
"first": "Joey",
"middle": [],
"last": "Faulkner",
"suffix": ""
},
{
"first": "Nick",
"middle": [],
"last": "Pawlowski",
"suffix": ""
},
{
"first": "Alan",
"middle": [],
"last": "Nichol",
"suffix": ""
}
],
"year": 2017,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Tom Bocklisch, Joey Faulkner, Nick Pawlowski, and Alan Nichol. 2017. Rasa: Open source language un- derstanding and dialogue management.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "Slugbot: Developing a computational model and framework of a novel dialogue genre",
"authors": [
{
"first": "Kevin",
"middle": [
"K"
],
"last": "Bowden",
"suffix": ""
},
{
"first": "Jiaqi",
"middle": [],
"last": "Wu",
"suffix": ""
},
{
"first": "Wen",
"middle": [],
"last": "Cui",
"suffix": ""
},
{
"first": "Juraj",
"middle": [],
"last": "Juraska",
"suffix": ""
},
{
"first": "Vrindavan",
"middle": [],
"last": "Harrison",
"suffix": ""
},
{
"first": "Brian",
"middle": [],
"last": "Schwarzmann",
"suffix": ""
},
{
"first": "Nick",
"middle": [],
"last": "Santer",
"suffix": ""
},
{
"first": "Marilyn",
"middle": [
"A"
],
"last": "Walker",
"suffix": ""
}
],
"year": 2019,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kevin K. Bowden, JiaQi Wu, Wen Cui, Juraj Juraska, Vrindavan Harrison, Brian Schwarzmann, Nick San- ter, and Marilyn A. Walker. 2019. Slugbot: Develop- ing a computational model and framework of a novel dialogue genre. CoRR, abs/1907.10658.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Deeppavlov: Open-source library for dialogue systems",
"authors": [
{
"first": "Mikhail",
"middle": [],
"last": "Burtsev",
"suffix": ""
},
{
"first": "Alexander",
"middle": [],
"last": "Seliverstov",
"suffix": ""
},
{
"first": "Rafael",
"middle": [],
"last": "Airapetyan",
"suffix": ""
},
{
"first": "Mikhail",
"middle": [],
"last": "Arkhipov",
"suffix": ""
},
{
"first": "Dilyara",
"middle": [],
"last": "Baymurzina",
"suffix": ""
},
{
"first": "Nickolay",
"middle": [],
"last": "Bushkov",
"suffix": ""
},
{
"first": "Olga",
"middle": [],
"last": "Gureenkova",
"suffix": ""
},
{
"first": "Taras",
"middle": [],
"last": "Khakhulin",
"suffix": ""
},
{
"first": "Yuri",
"middle": [],
"last": "Kuratov",
"suffix": ""
},
{
"first": "Denis",
"middle": [],
"last": "Kuznetsov",
"suffix": ""
}
],
"year": 2018,
"venue": "Proceedings of ACL 2018, System Demonstrations",
"volume": "",
"issue": "",
"pages": "122--127",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mikhail Burtsev, Alexander Seliverstov, Rafael Airapetyan, Mikhail Arkhipov, Dilyara Baymurz- ina, Nickolay Bushkov, Olga Gureenkova, Taras Khakhulin, Yuri Kuratov, Denis Kuznetsov, et al. 2018. Deeppavlov: Open-source library for dia- logue systems. In Proceedings of ACL 2018, System Demonstrations, pages 122-127.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "An integrated development environment for spoken dialogue systems",
"authors": [
{
"first": "Matthias",
"middle": [],
"last": "Denecke",
"suffix": ""
}
],
"year": 2000,
"venue": "Proceedings of the COLING-2000 Workshop on Using Toolsets and Architectures To Build NLP Systems",
"volume": "",
"issue": "",
"pages": "51--60",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Matthias Denecke. 2000. An integrated development environment for spoken dialogue systems. In Pro- ceedings of the COLING-2000 Workshop on Using Toolsets and Architectures To Build NLP Systems, pages 51-60, Centre Universitaire, Luxembourg. In- ternational Committee on Computational Linguis- tics.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Draw.io vs code integration",
"authors": [
{
"first": "Henning",
"middle": [],
"last": "Dieterichs",
"suffix": ""
},
{
"first": "Vincent",
"middle": [],
"last": "Rouill\u00e9",
"suffix": ""
}
],
"year": 2021,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Henning Dieterichs and Vincent Rouill\u00e9. 2021. Draw.io vs code integration. https: //github.com/hediet/vscode-drawio.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Analysing casual conversation",
"authors": [
{
"first": "S",
"middle": [],
"last": "Eggins",
"suffix": ""
},
{
"first": "D",
"middle": [],
"last": "Slade",
"suffix": ""
}
],
"year": 1996,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "S. Eggins and D. Slade. 1996. Analysing casual con- versation.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Emora stdm: A versatile framework for innovative dialogue system development",
"authors": [
{
"first": "James",
"middle": [
"D"
],
"last": "Finch",
"suffix": ""
},
{
"first": "Jinho",
"middle": [
"D"
],
"last": "Choi",
"suffix": ""
}
],
"year": 2020,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "James D. Finch and Jinho D. Choi. 2020. Emora stdm: A versatile framework for innovative dialogue sys- tem development.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Arindam Mandal, and Dilek Z. Hakkani-T\u00fcr. 2020. Further advances in open domain dialog systems in the third alexa prize socialbot grand challenge",
"authors": [
{
"first": "Raefer",
"middle": [],
"last": "Gabriel",
"suffix": ""
},
{
"first": "Yang",
"middle": [],
"last": "Liu",
"suffix": ""
},
{
"first": "Anna",
"middle": [],
"last": "Gottardi",
"suffix": ""
},
{
"first": "Mihail",
"middle": [],
"last": "Eric",
"suffix": ""
},
{
"first": "Anju",
"middle": [],
"last": "Khatri",
"suffix": ""
},
{
"first": "Anjali",
"middle": [],
"last": "Chadha",
"suffix": ""
},
{
"first": "Qinlang",
"middle": [],
"last": "Chen",
"suffix": ""
},
{
"first": "Behnam",
"middle": [],
"last": "Hedayatnia",
"suffix": ""
},
{
"first": "Pankaj",
"middle": [],
"last": "Rajan",
"suffix": ""
},
{
"first": "Ali",
"middle": [],
"last": "Binici",
"suffix": ""
},
{
"first": "Shui",
"middle": [],
"last": "Hu",
"suffix": ""
},
{
"first": "Karthik",
"middle": [],
"last": "Gopalakrishnan",
"suffix": ""
},
{
"first": "Seokhwan",
"middle": [],
"last": "Kim",
"suffix": ""
}
],
"year": null,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Raefer Gabriel, Yang Liu, Anna Gottardi, Mihail Eric, Anju Khatri, Anjali Chadha, Qinlang Chen, Behnam Hedayatnia, Pankaj Rajan, Ali Binici, Shui Hu, Karthik Gopalakrishnan, Seokhwan Kim, Lauren Stubel, Kate Bland, Arindam Mandal, and Dilek Z. Hakkani-T\u00fcr. 2020. Further advances in open do- main dialog systems in the third alexa prize socialbot grand challenge.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "An introduction to functional grammar",
"authors": [
{
"first": "M",
"middle": [
"A K"
],
"last": "Halliday",
"suffix": ""
},
{
"first": "Christian",
"middle": [],
"last": "Matthiessen",
"suffix": ""
},
{
"first": "M",
"middle": [
"I M"
],
"last": "",
"suffix": ""
}
],
"year": 2004,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "M. A. K. Halliday and Christian Matthiessen M. I. M. 2004. An introduction to functional grammar / M.A.K. Halliday, 3rd ed. / rev. by christian m.i.m. matthiessen. edition. Hodder Arnold London.",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "Arindam Mandal, Reza Ghanadan, and Dilek Hakkani-Tur. 2021. Further advances in open domain dialog systems in the fourth alexa prize socialbot grand challenge",
"authors": [
{
"first": "Shui",
"middle": [],
"last": "Hu",
"suffix": ""
},
{
"first": "Yang",
"middle": [],
"last": "Liu",
"suffix": ""
},
{
"first": "Anna",
"middle": [],
"last": "Gottardi",
"suffix": ""
},
{
"first": "Behnam",
"middle": [],
"last": "Hedayatnia",
"suffix": ""
},
{
"first": "Anju",
"middle": [],
"last": "Khatri",
"suffix": ""
},
{
"first": "Anjali",
"middle": [],
"last": "Chadha",
"suffix": ""
},
{
"first": "Qinlang",
"middle": [],
"last": "Chen",
"suffix": ""
},
{
"first": "Pankaj",
"middle": [],
"last": "Rajan",
"suffix": ""
},
{
"first": "Ali",
"middle": [],
"last": "Binici",
"suffix": ""
},
{
"first": "Varun",
"middle": [],
"last": "Somani",
"suffix": ""
},
{
"first": "Yao",
"middle": [],
"last": "Lu",
"suffix": ""
},
{
"first": "Prerna",
"middle": [],
"last": "Dwivedi",
"suffix": ""
},
{
"first": "Lucy",
"middle": [],
"last": "Hu",
"suffix": ""
},
{
"first": "Hangjie",
"middle": [],
"last": "Shi",
"suffix": ""
},
{
"first": "Sattvik",
"middle": [],
"last": "Sahai",
"suffix": ""
},
{
"first": "Mihail",
"middle": [],
"last": "Eric",
"suffix": ""
},
{
"first": "Karthik",
"middle": [],
"last": "Gopalakrishnan",
"suffix": ""
},
{
"first": "Seokhwan",
"middle": [],
"last": "Kim",
"suffix": ""
},
{
"first": "Spandana",
"middle": [],
"last": "Gella",
"suffix": ""
},
{
"first": "Alexandros",
"middle": [],
"last": "Papangelis",
"suffix": ""
},
{
"first": "Patrick",
"middle": [],
"last": "Lange",
"suffix": ""
},
{
"first": "Di Jin Nicole",
"middle": [],
"last": "Chartier",
"suffix": ""
},
{
"first": "Mahdi",
"middle": [],
"last": "Namazifar",
"suffix": ""
},
{
"first": "Aishwarya",
"middle": [],
"last": "Padmakumar",
"suffix": ""
},
{
"first": "Sarik",
"middle": [],
"last": "Ghazarian",
"suffix": ""
},
{
"first": "Shereen",
"middle": [],
"last": "Oraby",
"suffix": ""
},
{
"first": "Anjali",
"middle": [],
"last": "Narayan-Chen",
"suffix": ""
},
{
"first": "Yuheng",
"middle": [],
"last": "Du",
"suffix": ""
}
],
"year": null,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Shui Hu, Yang Liu, Anna Gottardi, Behnam He- dayatnia, Anju Khatri, Anjali Chadha, Qinlang Chen, Pankaj Rajan, Ali Binici, Varun Somani, Yao Lu, Prerna Dwivedi, Lucy Hu, Hangjie Shi, Sattvik Sahai, Mihail Eric, Karthik Gopalakrishnan, Seokhwan Kim, Spandana Gella, Alexandros Papangelis, Patrick Lange, Di Jin Nicole Chartier, Mahdi Namazifar, Aishwarya Padmakumar, Sarik Ghazarian, Shereen Oraby, Anjali Narayan-Chen, Yuheng Du, Lauren Stubell, Savanna Stiff, Kate Bland, Arindam Mandal, Reza Ghanadan, and Dilek Hakkani-Tur. 2021. Further advances in open do- main dialog systems in the fourth alexa prize social- bot grand challenge. https://d7qzviu3xw2xc. cloudfront.net/alexa/alexaprize/docs/ sgc4/Alexa-Prize-Technical-Paper-2021_ Final.pdf.",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "Speech and language processing : an introduction to natural language processing, computational linguistics, and speech recognition",
"authors": [
{
"first": "Dan",
"middle": [],
"last": "Jurafsky",
"suffix": ""
},
{
"first": "James",
"middle": [
"H"
],
"last": "Martin",
"suffix": ""
}
],
"year": 2009,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Dan Jurafsky and James H. Martin. 2009. Speech and language processing : an introduction to natural language processing, computational linguistics, and speech recognition. Pearson Prentice Hall, Upper Saddle River, N.J.",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "Alquist 4.0: Towards social intelligence using generative models and dialogue personalization",
"authors": [
{
"first": "Jakub",
"middle": [],
"last": "Konr\u00e1d",
"suffix": ""
},
{
"first": "Jan",
"middle": [],
"last": "Pichl",
"suffix": ""
},
{
"first": "Petr",
"middle": [],
"last": "Marek",
"suffix": ""
},
{
"first": "Petr",
"middle": [],
"last": "Lorenc",
"suffix": ""
},
{
"first": "Ondrej",
"middle": [],
"last": "Van Duy Ta",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Kobza",
"suffix": ""
}
],
"year": 2021,
"venue": "4th Proceedings of Alexa Prize",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Jakub Konr\u00e1d, Jan Pichl, Petr Marek, Petr Lorenc, Van Duy Ta, and Ondrej Kobza. 2021. Alquist 4.0: Towards social intelligence using generative models and dialogue personalization. 4th Proceedings of Alexa Prize.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "Dream technical report for the alexa prize",
"authors": [
{
"first": "Yuri",
"middle": [],
"last": "Kuratov",
"suffix": ""
},
{
"first": "Idris",
"middle": [],
"last": "Yusupov",
"suffix": ""
},
{
"first": "Dilyara",
"middle": [],
"last": "Baymurzina",
"suffix": ""
},
{
"first": "Denis",
"middle": [],
"last": "Kuznetsov",
"suffix": ""
},
{
"first": "Daniil",
"middle": [],
"last": "Cherniavskii",
"suffix": ""
},
{
"first": "Alexander",
"middle": [],
"last": "Dmitrievskiy",
"suffix": ""
},
{
"first": "Elena",
"middle": [],
"last": "Ermakova",
"suffix": ""
},
{
"first": "Fedor",
"middle": [],
"last": "Ignatov",
"suffix": ""
},
{
"first": "Dmitry",
"middle": [],
"last": "Karpov",
"suffix": ""
},
{
"first": "Daniel",
"middle": [],
"last": "Kornev",
"suffix": ""
}
],
"year": 2019,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Yuri Kuratov, Idris Yusupov, Dilyara Baymurzina, Denis Kuznetsov, Daniil Cherniavskii, Alexan- der Dmitrievskiy, Elena Ermakova, Fedor Ignatov, Dmitry Karpov, Daniel Kornev, et al. 2019. Dream technical report for the alexa prize 2019. 3rd Pro- ceedings of Alexa Prize.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "Dialog flow framework @ deepmipt github",
"authors": [
{
"first": "Denis",
"middle": [],
"last": "Kuznetsov",
"suffix": ""
},
{
"first": "Dmitry",
"middle": [],
"last": "Evseev",
"suffix": ""
}
],
"year": 2021,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Denis Kuznetsov and Dmitry Evseev. 2021. Di- alog flow framework @ deepmipt github. https://github.com/deepmipt/dialog_ flow_framework/.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Visual skill building",
"authors": [
{
"first": "Andrew",
"middle": [],
"last": "Lawrence",
"suffix": ""
},
{
"first": "Braden",
"middle": [],
"last": "Ream",
"suffix": ""
},
{
"first": "Michael",
"middle": [],
"last": "Hood",
"suffix": ""
},
{
"first": "Tyler",
"middle": [],
"last": "Han",
"suffix": ""
}
],
"year": 2021,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Andrew Lawrence, Braden Ream, Michael Hood, and Tyler Han. 2021. Visual skill building. https:// www.voiceflow.com/.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Chat and goal-oriented dialog together: a unified example-based architecture for multi-domain dialog management",
"authors": [
{
"first": "Cheongjae",
"middle": [],
"last": "Lee",
"suffix": ""
},
{
"first": "Sangkeun",
"middle": [],
"last": "Jung",
"suffix": ""
},
{
"first": "Minwoo",
"middle": [],
"last": "Jeong",
"suffix": ""
},
{
"first": "Gary Geunbae",
"middle": [],
"last": "Lee",
"suffix": ""
}
],
"year": 2006,
"venue": "IEEE Spoken Language Technology Workshop",
"volume": "",
"issue": "",
"pages": "194--197",
"other_ids": {
"DOI": [
"10.1109/SLT.2006.326788"
]
},
"num": null,
"urls": [],
"raw_text": "Cheongjae Lee, Sangkeun Jung, Minwoo Jeong, and Gary Geunbae Lee. 2006. Chat and goal-oriented di- alog together: a unified example-based architecture for multi-domain dialog management. In 2006 IEEE Spoken Language Technology Workshop, pages 194- 197.",
"links": null
},
"BIBREF20": {
"ref_id": "b20",
"title": "Opendial: A toolkit for developing spoken dialogue systems with probabilistic rules",
"authors": [
{
"first": "Pierre",
"middle": [],
"last": "Lison",
"suffix": ""
},
{
"first": "Casey",
"middle": [],
"last": "Kennington",
"suffix": ""
}
],
"year": 2016,
"venue": "Proceedings of ACL-2016 system demonstrations",
"volume": "",
"issue": "",
"pages": "67--72",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Pierre Lison and Casey Kennington. 2016. Opendial: A toolkit for developing spoken dialogue systems with probabilistic rules. In Proceedings of ACL- 2016 system demonstrations, pages 67-72.",
"links": null
},
"BIBREF21": {
"ref_id": "b21",
"title": "Getting to know each other: The role of social dialogue in recovery from errors in social robots",
"authors": [
{
"first": "M",
"middle": [],
"last": "Gale",
"suffix": ""
},
{
"first": "Jill",
"middle": [],
"last": "Lucas",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Boberg",
"suffix": ""
},
{
"first": "Ron",
"middle": [],
"last": "Traum",
"suffix": ""
},
{
"first": "Jonathan",
"middle": [],
"last": "Artstein",
"suffix": ""
},
{
"first": "Alesia",
"middle": [],
"last": "Gratch",
"suffix": ""
},
{
"first": "Emmanuel",
"middle": [],
"last": "Gainer",
"suffix": ""
},
{
"first": "Anton",
"middle": [],
"last": "Johnson",
"suffix": ""
},
{
"first": "Mikio",
"middle": [],
"last": "Leuski",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Nakano",
"suffix": ""
}
],
"year": 2018,
"venue": "Proceedings of the 2018 ACM/IEEE International Conference on Human-Robot Interaction, HRI '18",
"volume": "",
"issue": "",
"pages": "344--351",
"other_ids": {
"DOI": [
"10.1145/3171221.3171258"
]
},
"num": null,
"urls": [],
"raw_text": "Gale M. Lucas, Jill Boberg, David Traum, Ron Art- stein, Jonathan Gratch, Alesia Gainer, Emmanuel Johnson, Anton Leuski, and Mikio Nakano. 2018. Getting to know each other: The role of social di- alogue in recovery from errors in social robots. In Proceedings of the 2018 ACM/IEEE International Conference on Human-Robot Interaction, HRI '18, page 344-351, New York, NY, USA. Association for Computing Machinery.",
"links": null
},
"BIBREF22": {
"ref_id": "b22",
"title": "Experiments testing gpt-3's ability at commonsense reasoning: results",
"authors": [
{
"first": "Gary",
"middle": [],
"last": "Marcus",
"suffix": ""
},
{
"first": "Ernest",
"middle": [],
"last": "Davis",
"suffix": ""
}
],
"year": 2020,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Gary Marcus and Ernest Davis. 2020. Experiments testing gpt-3's ability at commonsense reasoning: re- sults. https://cs.nyu.edu/~davise/papers/ GPT3CompleteTests.html.",
"links": null
},
"BIBREF23": {
"ref_id": "b23",
"title": "The Penn Discourse TreeBank 2.0",
"authors": [
{
"first": "Rashmi",
"middle": [],
"last": "Prasad",
"suffix": ""
},
{
"first": "Nikhil",
"middle": [],
"last": "Dinesh",
"suffix": ""
},
{
"first": "Alan",
"middle": [],
"last": "Lee",
"suffix": ""
},
{
"first": "Eleni",
"middle": [],
"last": "Miltsakaki",
"suffix": ""
},
{
"first": "Livio",
"middle": [],
"last": "Robaldo",
"suffix": ""
},
{
"first": "Aravind",
"middle": [],
"last": "Joshi",
"suffix": ""
},
{
"first": "Bonnie",
"middle": [],
"last": "Webber",
"suffix": ""
}
],
"year": 2008,
"venue": "Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08)",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Rashmi Prasad, Nikhil Dinesh, Alan Lee, Eleni Milt- sakaki, Livio Robaldo, Aravind Joshi, and Bon- nie Webber. 2008. The Penn Discourse TreeBank 2.0. In Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08), Marrakech, Morocco. European Lan- guage Resources Association (ELRA).",
"links": null
},
"BIBREF24": {
"ref_id": "b24",
"title": "Sk Jayadevan, Gene Hwang, and Art Pettigrue",
"authors": [
{
"first": "Ashwin",
"middle": [],
"last": "Ram",
"suffix": ""
},
{
"first": "Rohit",
"middle": [],
"last": "Prasad",
"suffix": ""
},
{
"first": "Chandra",
"middle": [],
"last": "Khatri",
"suffix": ""
},
{
"first": "Anu",
"middle": [],
"last": "Venkatesh",
"suffix": ""
},
{
"first": "Raefer",
"middle": [],
"last": "Gabriel",
"suffix": ""
},
{
"first": "Qing",
"middle": [],
"last": "Liu",
"suffix": ""
},
{
"first": "Jeff",
"middle": [],
"last": "Nunn",
"suffix": ""
},
{
"first": "Behnam",
"middle": [],
"last": "Hedayatnia",
"suffix": ""
},
{
"first": "Ming",
"middle": [],
"last": "Cheng",
"suffix": ""
},
{
"first": "Ashish",
"middle": [],
"last": "Nagar",
"suffix": ""
},
{
"first": "Eric",
"middle": [],
"last": "King",
"suffix": ""
}
],
"year": 2018,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ashwin Ram, Rohit Prasad, Chandra Khatri, Anu Venkatesh, Raefer Gabriel, Qing Liu, Jeff Nunn, Behnam Hedayatnia, Ming Cheng, Ashish Nagar, Eric King, Kate Bland, Amanda Wartick, Yi Pan, Han Song, Sk Jayadevan, Gene Hwang, and Art Pet- tigrue. 2018. Conversational ai: The science behind the alexa prize.",
"links": null
},
"BIBREF26": {
"ref_id": "b26",
"title": "Ddidde alpha version 0.2.1",
"authors": [
{
"first": "Oleg",
"middle": [],
"last": "Serikov",
"suffix": ""
},
{
"first": "Dmitry",
"middle": [],
"last": "Babadeev",
"suffix": ""
}
],
"year": 2021,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Oleg Serikov and Dmitry Babadeev. 2021. Dd- idde alpha version 0.2.1. https://github.com/",
"links": null
},
"BIBREF27": {
"ref_id": "b27",
"title": "Number of digital voice assistants in use worldwide from",
"authors": [
{
"first": "",
"middle": [],
"last": "Statista",
"suffix": ""
}
],
"year": 2019,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Statista. 2020. Number of digital voice assistants in use worldwide from 2019 to 2024 (in billions)*. https: //www.statista.com/statistics/973815/ worldwide-digital-voice-assistant-in-use/.",
"links": null
},
"BIBREF28": {
"ref_id": "b28",
"title": "Pydial: A multidomain statistical dialogue system toolkit",
"authors": [
{
"first": "Stefan",
"middle": [],
"last": "Ultes",
"suffix": ""
},
{
"first": "Lina M Rojas",
"middle": [],
"last": "Barahona",
"suffix": ""
},
{
"first": "Pei-Hao",
"middle": [],
"last": "Su",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Vandyke",
"suffix": ""
},
{
"first": "Dongho",
"middle": [],
"last": "Kim",
"suffix": ""
},
{
"first": "Inigo",
"middle": [],
"last": "Casanueva",
"suffix": ""
},
{
"first": "Pawe\u0142",
"middle": [],
"last": "Budzianowski",
"suffix": ""
},
{
"first": "Nikola",
"middle": [],
"last": "Mrk\u0161i\u0107",
"suffix": ""
},
{
"first": "Tsung-Hsien",
"middle": [],
"last": "Wen",
"suffix": ""
},
{
"first": "Milica",
"middle": [],
"last": "Gasic",
"suffix": ""
}
],
"year": 2017,
"venue": "Proceedings of ACL 2017, System Demonstrations",
"volume": "",
"issue": "",
"pages": "73--78",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Stefan Ultes, Lina M Rojas Barahona, Pei-Hao Su, David Vandyke, Dongho Kim, Inigo Casanueva, Pawe\u0142 Budzianowski, Nikola Mrk\u0161i\u0107, Tsung-Hsien Wen, Milica Gasic, et al. 2017. Pydial: A multi- domain statistical dialogue system toolkit. In Pro- ceedings of ACL 2017, System Demonstrations, pages 73-78.",
"links": null
},
"BIBREF29": {
"ref_id": "b29",
"title": "Chatbot, natural language processing popularity",
"authors": [],
"year": 2021,
"venue": "",
"volume": "2",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Google Books Ngram Viewer. 2021. Chat- bot, natural language processing popularity. https://books.google.com/ngrams/graph? content=chatbot%2C+natural+language+ processing&year_start=2000.",
"links": null
},
"BIBREF30": {
"ref_id": "b30",
"title": "Artificial intelligence markup language (aiml)",
"authors": [
{
"first": "Richard",
"middle": [],
"last": "Wallace",
"suffix": ""
}
],
"year": 2001,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Richard Wallace. 2001. Artificial intelligence markup language (aiml). https://web. archive.org/web/20060505035935/http: //aitools.org/aiml/spec/.",
"links": null
},
"BIBREF31": {
"ref_id": "b31",
"title": "The Anatomy of",
"authors": [
{
"first": "Richard",
"middle": [
"S"
],
"last": "Wallace",
"suffix": ""
}
],
"year": 2009,
"venue": "",
"volume": "",
"issue": "",
"pages": "181--210",
"other_ids": {
"DOI": [
"10.1007/978-1-4020-6710-5_13"
]
},
"num": null,
"urls": [],
"raw_text": "Richard S. Wallace. 2009. The Anatomy of A.L.I.C.E., pages 181-210. Springer Netherlands, Dordrecht.",
"links": null
},
"BIBREF32": {
"ref_id": "b32",
"title": "Gunrock: A social bot for complex and engaging long conversations",
"authors": [
{
"first": "Dian",
"middle": [],
"last": "Yu",
"suffix": ""
},
{
"first": "Michelle",
"middle": [],
"last": "Cohn",
"suffix": ""
},
{
"first": "Yi",
"middle": [
"Mang"
],
"last": "Yang",
"suffix": ""
},
{
"first": "Chun-Yen",
"middle": [],
"last": "Chen",
"suffix": ""
},
{
"first": "Weiming",
"middle": [],
"last": "Wen",
"suffix": ""
},
{
"first": "Jiaping",
"middle": [],
"last": "Zhang",
"suffix": ""
},
{
"first": "Mingyang",
"middle": [],
"last": "Zhou",
"suffix": ""
}
],
"year": 2019,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Dian Yu, Michelle Cohn, Yi Mang Yang, Chun-Yen Chen, Weiming Wen, Jiaping Zhang, Mingyang Zhou, Kevin Jesse, Austin Chau, Antara Bhowmick, Shreenath Iyer, Giritheja Sreenivasulu, Sam David- son, Ashwin Bhandare, and Zhou Yu. 2019. Gun- rock: A social bot for complex and engaging long conversations.",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"text": "New .drawio file created in VS Code Figure 2: Setting Speech Function for the current node",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF1": {
"text": "Using the Speech Function Recommendations",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF2": {
"text": "Outlining Scenario In VS Code ExtensionFigure 5: Using the Show Markup functionality to generate DFF Markup Code Figure 6: Adding Custom Code to the DFF Markup File",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF3": {
"text": "Statistic data frame is shown in the Dashboard Transition graph with counters is built for all dialogues. Counters show numbers of users hitting these nodes or edges.",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF4": {
"text": "These figures are built in the dashboard. The top figure shows nodes that will be sent by all users at a certain step of the dialogue. The abscissa axis shows the order of a step in the dialog, the ordinate axis shows the node in which the user is located for this step. The bottom figure shows the frequency of users hitting each node, different colors are shown in different flows in which the nodesFigure 10: Statistic of transitions in dashboard. The top figure shows the frequency of users hitting each transition between nodes. The bottom figure shows duration of processing time for each transition between nodes. Different colors are shown in different flows in which the nodes.",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF5": {
"text": "Statistic of transition in dashboard for one dialog. This figure shows the user transitions from node to node.",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF6": {
"text": "This figure is built in the dashboard. The graph shows user transitions through the script graph using one dialog as an example. Counters show numbers of users hitting these nodes or edges.",
"uris": null,
"type_str": "figure",
"num": null
},
"FIGREF7": {
"text": "Example of dialog flow in DFF DSL format Figure 14: DFF DSL elements for entity extraction from user utterance and slot filling in response Figure 15: DFF DSL elements for entity extraction using Entity Detection annotator Figure 16: Dialog Flow with \"fact_provider\" -a function for response generation using a wikiHow article Figure 17: Dialogflow for generic responses generationFigure 18: Coronavirus skill in pure Python Figure 19: Coronavirus skill DFF DSL format",
"uris": null,
"type_str": "figure",
"num": null
}
}
}
} |