Abstract
In the face of the information recommendation requirements in mobile Internet applications, in order to better use the user micro implicit feedback behavior obtained by the mobile intelligent terminal to improve the recommendation efficiency, this paper intends to carry out the analysis of the implicit feedback behavior by analyzing the behavior distribution and behavior correlation. The analytical results reveal the particularity of the implicit feedback behavior in mobile intelligent terminal.
You have full access to this open access chapter, Download conference paper PDF
Similar content being viewed by others
Keywords
1 Introduction
The analysis of user network behavior characteristics is the design basis of many Internet products. Through in-depth analysis of user behavior, completing personalized recommendation can bring users a better application experience. In the field of market-driven software engineering, user behavior analysis also provides new ideas and improvement direction for application development to meet the requirements of the new situation.
User network behavior can be divided into two categories: explicit feedback behavior and implicit feedback behavior. The definition, characteristics, differences and types of the two types of behavior, relatively stable and unified views have been formed. Display feedback behavioral data can accurately express user intention, but because it interferes with the normal interaction process in the network, increases the cognitive burden and reduces the user experience, it is difficult to obtain data. On the contrary, for users’ implicit feedback behavior data, it is much less difficult to obtain and has large information abundance. Therefore, although such information has low accuracy, large data noise and large context sensitivity, this research field is still getting more and more attention.
2 Related Studies
With the rapid development of social networks and e-commerce, the number of Internet users has increased and the demand for personalized recommendation services is growing. It is the focus and difficulty of current research to deal with the massive amount of multi-source heterogeneous data generated when users browse the mobile Internet.
The original personalized recommendation service is mainly for PC-based users, and the relevant research is mainly divided into the following four aspects: research on an application scenario, a kind or technology, recommendation system evaluation method, and a kind of common problems in the recommendation system.
The study of user network behavior was initially applied in the field of information retrieval, which significantly improves the performance of information filtering compared to other feedback, and quickly filters from massive information sets, providing the retrieval set with the highest correlation with their interest preferences[1]. Lots of researches show that user browsing time is important to find person’s preference [2, 3]. Moreover, bookmarking, printing and saving could show users’ interesting. Oard and Kim clustered them into three groups [4,5,6].
In addition, mobile network environment give a challenge. Researches such as [7, 8] focus on this condition. Implicit behaviors from user exploring website in this condition are hot [9,10,11]. Therefore, this paper conducts the analysis of the implicit feedback behavior of mobile intelligent terminals.
3 Problem Description and Behavioral Analysis
3.1 Problem Description
Users’ network implicit behavior contains information about their preferences, but it is generally not clearly expressed, so it is more difficult to correctly judge their preferences, and the researchers have carried out more work in this regard. At present, there are many implicit studies on macro-network behavior, such as behavioral sequence analysis or item recommendation based on browsing, adding shopping carts, buying and other behaviors. For the implicit feedback behavior of user micro network, there are few studies and conclusions that are found due to small data scale, less data category and low data dimension. This paper plans to carry out implicit feedback behavior analysis, explore the characteristics of implicit feedback behavior data, and lay the foundation for the subsequent recommendation based on implicit feedback behavior.
3.2 Analysis of User Microscopic Implicit Feedback Behavior
Acquiring approach of users’ micro implicit behavior includes two ways. The first one is direct acquiring way, which is conducted by running some software in background. The other is indirect way, generally speaking, which is acquired by questionnaire. In direct acquisition, there are problems of sparse data, less categories and low dimensions, which is not conducive to subsequent analysis and deterministic conclusions. In this paper, we use data in indirect acquisition mode to analyze the micro indirect feedback behavior, extracting part of the survey content (Q4-Q15) from the questionnaire, and mapping it to micro implicit behaviors, IFBn above, from user exploring in website, as below in Table 1.
In order to facilitate the subsequent association analysis of various kinds of influence variables, the user micro implicit feedback behavior is divided into two categories according to the questionnaire data: 1) mutually exclusive micro implicit feedback behavior and 2) non-mutually exclusive micro implicit feedback behavior in the literature [7]. Among them, IFB1-IFB3 is category mutually exclusive micro implicit feedback behavior, each user corresponds to a micro implicit feedback behavior result, such as selecting only one application market class, a certain access frequency and attention frequency to item determined; IFB4-IFB11 is category non-mutually exclusive type micro implicit feedback behavior, each user can correspond to multiple micro implicit feedback behavior results, such as the query frequency to item when the user is depressed, when the user needs to complete the task, when the user is bored.
The variable \(f_{{\text{IFBn}}} (C_m )\) is defined as the occurrence frequency of some implicit behavior IFBn. Then for mutually exclusive user behavior, \(f_{{\text{IFBn}}} (C_m ) = \sum_m {C_m } = 1\), and for non-mutually exclusive user behavior, \(f_{{\text{IFBn}}} (C_m ) = \sum_m {C_m } \ge 1\). Among these, \(C_m\) is the \(m^{{\text{th}}}\) the category attribute values of the \(n^{{\text{th}}}\) micro implicit feedback behavior IFBn.
Let the sample size of user micro implicit feedback behavior be \(N\), then the behavior distribution is defined as \({{\left( {\sum_N {Cm} } \right)} / N}\) to clearly reflect the differences of various attributes of user micro implicit feedback behavior. At the same time, the correlation of the behavior by calculating the micro implicit feedback behavior. Due to the large numerical discretization, \(f_{{\text{IFBn}}} (C_m )\), of the microscopic implicit feedback behavior IFBn and the inconsistent range of variation, it was normalized before the correlation analysis.
4 Experiments and Analysis
4.1 Microscopic Implicit Feedback Behavior Distribution
-
1)
Users differ greatly in category selection (IFB1) for the application market. In Fig. 1, the top three are the differences in micro implicit feedback behavior of Android Market, Apple iOS App Store, Nokia Ovi Store, except from the context influence of user attributes discussed here, and more from the influence of software and hardware of mobile intelligent terminals, which will be discussed in subsequent studies.
-
2)
The frequency of access (IFB2) in the application market is the reflection of user demand. This statistical data has not a strong relationship between the hardware and software of the mobile intelligent terminals used by the user, so the category is relatively evenly distributed, as shown in Fig. 2.
-
3)
The number of attention to item per month (IFB3) reflects the strong willingness and choice tendency, but few users with high attention, as shown in Fig. 3, more users pay attention to item within 5 times a month, among which the number of attention to item is 0 or 1 is 40% and 2–5 for 36%, showing certain long tail characteristics.
-
4)
The query frequency (IFB4) to item is also a microscopic implicit feedback behavior that reflects user willingness and choice propensity. According to the questionnaire data of literature [7], except for the last category (including data that cannot be classified to the top 5 categories), users with different needs, such as work demand, query demand, entertainment demand, etc., the query frequency fluctuates little, as shown in Fig. 4.
-
5)
The query way of item (IFB5), from the questionnaire data in the literature [7], except the last category (including data that cannot be categorized to the top 8 categories), is shown in Fig. 5. the most way users use to query of item is keyword search, the most distrust way is list ranking.
-
6)
Detail level of item browsing (IFB6). The most user attention to item information is price, features, detail description and comments, as shown in Fig. 6. From the implicit feedback behavior of mobile smart terminals, it is similar to PC-based user behavior.
-
7)
The intensity of attention on item (IFB7) also reflects user purchase possibilities for item. In addition to the last category (including data that cannot be classified to the top 14 categories), item with high intensity of user attention are entertainment, function and novelty, and lower ones are stranger communication, advertising effect and impulse purchase, reflecting users’ rational attention, as shown in Fig. 7.
-
8)
Purchases of item (IFB8). Except for the last category (including data that cannot be categorized to the top 11 categories), users preferred free item, unless there is no free version and similar features and requires increased functionality and performance, as shown in Fig. 8. Users don’t tend to subscribe to a certain item and pay.
-
9)
Evaluation behavior (IFB9) for item. Except for the last category (including data that cannot be categorized to the top 6 categories), the data showed that the user did not like the evaluation, as shown in Fig. 9. Some existing reviews are given mainly to let others understand the merits of item. Mandatory evaluations are currently relatively few.
-
10)
Cancel attention to item (IFB10). Except for the last category (including data that cannot be classified to the top 14 categories), causes users to dismiss item or find a better replacement, as shown in Fig. 10. The cancellation of attention is less affected by his family or friends.
-
11)
Category focus behavior on item (IFB11). In addition to the last category (including data that cannot be classified to the top 22 categories), the item categories that users focus on are game category, social network category, music category, etc., and the item categories that users do not pay attention to are catalog category, medicine category and reference category, as shown in Fig. 11.
4.2 Microscopic Implicit Feedback Behavioral Correlations
The correlations between the implicit feedback behavior of non-mutually exclusive type microscopy are analyzed, as shown in Table 2.
The significance value indicators in the table are all 0, less than 0.05, meeting the premise of correlation analysis.The Pearson correlation value of IFB4 with IFB 5, IFB 7 was greater than 0.6, indicating that the three microscopic implicit feedback behaviors are correlated and strongly correlated. Similarly, IFB 5 is associated strongly with IFB 6, IFB 7, IFB 6 with IFB 7, IFB 10, and IFB 7 with IFB 11. Purchase behavior (IFB8) for item and evaluation behavior for item (IFB9), showed a weak correlation with other behaviors.
5 Conclusions
This paper provides the analysis of the implicit feedback behavior of mobile intelligent terminal, establishes the micro implicit feedback behavior data set, and analyzes the behavior distribution and non-mutually exclusive micro implicit feedback behavior respectively, which lays the basis for further using the analysis results.
References
Seo, Y.W., Zhang, B.T.: Learning user’s preferences by analyzing Webbrowsing behaviors. In: Proceedings the 4th International Conference on Autonomous Agents, pp. 381–387 (2000)
Morita, M., Shinoda, Y.: Information filtering based on user behavior analysis and best match text retrieval. In: Proceedings the 17th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, pp. 272–281 (1994)
Konstan, J.A., Miller, B.N., Maltz, D., et al.: GroupLens: applying collaborative filtering to Usenet news. In: Communications of the ACM, pp. 77–87 (1997)
Oard, D.W., Kim, J.: Implicit feedback for recommender systems. In: Proceedings of the AAAI Workshop on Recommender Systems, p. 83 (1998)
Kelly, D., Teevan, J.: Implicit feedback for inferring user preference: a bibliography. In: Acm Sigir Forum, pp. 18–28 (2003)
Yin, C.H., Deng, W.: Extracting user interests based on analysis of user behaviors. Comput. Technol. Dev. 18(5), 37–39 (2008)
Lim, S.L., Bentley, P.J.: Investigating country differences in mobile App user behavior and challenges for software engineering. IEEE Trans. Softw. Eng. 41(1), 40–64 (2015)
Zhou, G., Zhu, X., Song, C., et al.: Deep interest network for click-through rate prediction. In: Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 1059–1068 (2018)
Xiao, Z., Yang, L., Jiang, W., et al.: Deep multi-interest network for click-through rate prediction. In: Proc. of the 29th ACM International Conference on Information & Knowledge Management, pp. 2265–2268 (2020)
Tang, H., Liu, J., Zhao, M., et al.: Progressive layered extraction (PLE): a novel multi-task learning (MTL) model for personalized recommendations. In: Fourteenth ACM Conference on Recommender Systems, pp. 269–278 (2020)
Qu, J.: Big data network user rowsing implicit feedback information retrieval simulation. Computer Simulation 430–433 (2019)
Acknowledgments
This work was supported by The National Natural Science Foundation of China (No. 61802107); Science and technology research project of Hebei University (No. ZD2020171); Jiangsu Planned Projects for Postdoctoral Research Funds (No. 1601085C).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Copyright information
© 2022 The Author(s)
About this paper
Cite this paper
Wang, W., Zhang, C., Zheng, X., Du, Y. (2022). Analysis of the Micro Implicit Feedback Behavior of User Network Exploring Based on Mobile Intelligent Terminal. In: Qian, Z., Jabbar, M., Li, X. (eds) Proceeding of 2021 International Conference on Wireless Communications, Networking and Applications. WCNA 2021. Lecture Notes in Electrical Engineering. Springer, Singapore. https://doi.org/10.1007/978-981-19-2456-9_4
Download citation
DOI: https://doi.org/10.1007/978-981-19-2456-9_4
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-19-2455-2
Online ISBN: 978-981-19-2456-9
eBook Packages: EngineeringEngineering (R0)