Abstract: 【Purpose/significance】Knowledge can be transferred, fermented, and deepened in the process of group social interaction,
and new knowledge can be generated. Exploring the social interaction quality of academic virtual communities will help optimize the
academic research knowledge service system, thereby improving the efficiency of knowledge exchange.【Method/process】The social
interaction quality evaluation index system is constructed according to the theory of remote interaction hierarchy model, and the mat‐
ter-element extension method was used to empirically test the social interaction quality level of the“Xiaomuchong”virtual communi‐
ty, and we also use LASSO regression model for sensitivity analysis.【Result/conclusion】The interaction quality of“Xiaomuchong”
has reached a good level overall, tendency to a general level, in which the quality of operational interaction and conceptual interaction
is good, while the quality of information interaction and emotional interaction is average. The second-level indicators reflect differenc‐
es. Sensitivity analysis verifies the rationality and effectiveness of the evaluation method, which opens up a new path for follow-up re‐
search【. Innovation/limitation】The sample data of case analysis is small, and the evaluation method of expert scoring has certain sub‐jectivity. The follow-up study will further expand the scope of the investigation to ensure the objectivity and universality of the re‐
search conclusions.