TY - GEN
T1 - Impression and Suitability of Q&A Statements through Factor Scores Using 2-Gram
AU - Yokoyama, Yuya
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - In order to solve the mismatches between the intentions of questioners and respondents at Question and Answer (Q&A) sites, nine factors of impressions were obtained through impression evaluation experiments. Through multiple regression analysis, factor scores were then estimated using feature values of statements. The factor scores obtained were subsequently used for finding the respondent who would be expected to appropriately answer a question. However, this method largely depended on the syntactic information extracted through morphological analysis. Thus, it was also partially shown that N-gram was also applicable as an alternative method. In applying N-gram, estimation of factor scores has been realized so far. Therefore, in this paper, using the distances between the factor scores of Q&A statements, whether applying N-gram would also be applicable to assessing impression and suitability for the purpose of detecting the respondent who can appropriately answer a question is inspected. As a first phase of using N-gram, in this analysis, 2-gram has been tentatively utilized. The analysis result using 2-gram has shown that a similar tendency as the previous analysis employing Syn-Info is observed. It has also been shown that users with sufficient knowledge in specific categories can be thought as appropriate respondents.
AB - In order to solve the mismatches between the intentions of questioners and respondents at Question and Answer (Q&A) sites, nine factors of impressions were obtained through impression evaluation experiments. Through multiple regression analysis, factor scores were then estimated using feature values of statements. The factor scores obtained were subsequently used for finding the respondent who would be expected to appropriately answer a question. However, this method largely depended on the syntactic information extracted through morphological analysis. Thus, it was also partially shown that N-gram was also applicable as an alternative method. In applying N-gram, estimation of factor scores has been realized so far. Therefore, in this paper, using the distances between the factor scores of Q&A statements, whether applying N-gram would also be applicable to assessing impression and suitability for the purpose of detecting the respondent who can appropriately answer a question is inspected. As a first phase of using N-gram, in this analysis, 2-gram has been tentatively utilized. The analysis result using 2-gram has shown that a similar tendency as the previous analysis employing Syn-Info is observed. It has also been shown that users with sufficient knowledge in specific categories can be thought as appropriate respondents.
KW - Euclidean distance
KW - factor score
KW - multiple regression analysis
KW - N-gram
KW - Q&A site
UR - https://www.scopus.com/pages/publications/85191506834
UR - https://www.scopus.com/pages/publications/85191506834#tab=citedBy
U2 - 10.1109/IIAI-AAI-Winter61682.2023.00017
DO - 10.1109/IIAI-AAI-Winter61682.2023.00017
M3 - Conference contribution
AN - SCOPUS:85191506834
T3 - Proceedings - 2023 15th International Congress on Advanced Applied Informatics Winter, IIAI-AAI-Winter 2023
SP - 45
EP - 51
BT - Proceedings - 2023 15th International Congress on Advanced Applied Informatics Winter, IIAI-AAI-Winter 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 15th International Congress on Advanced Applied Informatics Winter, IIAI-AAI-Winter 2023
Y2 - 11 December 2023 through 13 December 2023
ER -