| With the rapid development of digital technology and digital related industries,it also brings highly fierce competition in the mobile application market.In order to cope with this competition and better discover and meet user needs,APP software has to be updated frequently.Previous studies have pointed out that the way of product and service innovation relying solely on the internal knowledge of enterprises can no longer adapt to today’s highly competitive market environment,and the voice of users is an external resource with important value.Therefore,it is very important for enterprises to absorb and adopt the needs of users in the process of product and service innovation.Previous literatures have expounded the importance and necessity of user suggestions in product upgrading and development from the perspectives of antecedents,importance and paths.However,few scholars have discussed the attitude of APP towards user suggestions in the process of updating different types(function or operation level)and the extent to which user suggestions will be adopted under the background of mobile application market.Therefore,the first research question of this paper is: how to identify whether the app has adopted user suggestions in the iterative update process of APP software? Furthermore,this paper puts forward the second research question: what factors will affect the adoption of user suggestions by app?In order to answer the above questions,this paper discriminates the semantic correlation between the update log data of APP software and the user comment data,and studies the adoption of user suggestions in each update cycle of app;Based on the perspective of user participation,this paper discusses the factors that affect the adoption of user suggestions by app.In the part of identifying whether the app adopts user suggestions,this paper proposes a set of semantic fusion model based on encoder decoder,which maps the data from different data sources to the common semantic subspace by semantic transformation of the update log data and user comment data,and then obtains the joint representation of the two,and then discriminates their semantic correlation through the classifier,It can effectively distinguish the semantic correlation between the update log data and user comment data,and calculate the app’s adoption of user suggestions in each update cycle.Finally,the effectiveness of the model proposed in this paper is verified by the app’s update log data and user comment data in the real mobile application market,and then quantify the app’s adoption of user suggestions.The research results show that less than 40% of user suggestions are adopted by app for updates at the functional and operational levels.In the part of influencing factors of APP adoption of user suggestions,based on the perspective of user participation,this paper puts forward relevant assumptions from the perspective of user comments and user ratings,and tests the impact of the number of user comments and user ratings on app adoption of user suggestions.The study found that the number of user comments had an inverted U-shaped effect on the degree of APP adoption of user suggestions;User score and score variance have a positive impact on the degree of APP adoption of user suggestions.Through the above research,this paper has two contributions.First,this paper proposes a semantic fusion model of multi-source data based on encoder decoder.The model maps the data from different data sources to the same semantic subspace through encoder decoder structure,and then obtains the joint representation of the two,which can effectively solve the problem of semantic correlation between multi-source data such as update logs and user comments.Second,this paper studies the factors that affect the adoption of user suggestions in the process of APP software iterative update in the mobile application market,enriching the relevant research on the impact of online user comments on app software iterative update. |
PDF Full Text Request