| Information seeking is an important stage of consumers'purchase decision process, and online information seeking has long been consumers'primary motivation to use internet. Online information seeking can improve the consumers'online purchase intention and satisfaction, so it has become a research topic in e-commerce domain. The new characteristics of online information environment underline the importance such research.The paper perceives online information seeking as follows: the formation of information need shaped by the interaction of the consumer and the task context confronted the consumer; Information need will trigger and influence the information seeking behavior; Online information seeking can be divided into two basic processes, namely, visit on websites and transfer between websites. Based on the perception, we pay our attention mainly on: defining information need, exploring the formation mechanism and measurement of information need, proposing and validating consumers'information need model; analyzing the effect of information need on online information behavior, and comparing the effect of information need with that of other factors from competing models; and investigating how consumers allocate their search efforts and transfer among a wide range of websites. Our major achievements and initiatives are as follows:(1) Proposing and validating information need model. We defines information need as the needs for external information derived from the consumer's recognition that the owned knowledge is not convinced to be sufficient to solve the current purchase task confronted him. Then the hypotheses concerning the relation between information need and the antecedent factors, individual's prior knowledge and task context, and the measurements of these three variables (information need, prior knowledge and task context) are proposed. A information need model is formed based on these hypotheses and measurements. Finally, the information need model is validated against empirical data collected under quasi-experiment environment. The initiatives of the model lie in:①It is confirmed that both individual's prior knowledge and task context have significant influence on information need. Where individual's prior knowledge is measured by domain knowledge and internet knowledge, and task context is measured by task complexity and task risk;②The amount of information need and the clarity of information need are proposed for the measurement of information need, the amount dimension quantifies the static attribute of information need, while the clarity dimension characterizes the dynamic development of information need. And we further analyze the structure and development of information need and find:①The quantity and development of different types of information need present significant differences;②The development of information need is centered on information task or topic, which will lead consumers devote their attention to seek for information of a certain type or topic in a particular period;③Information need evolve along with problem solving model (from"Problem Statement"to"Problem Solving"). Consumers first seek for principle knowledge to deep their understanding of the problem, and then consumers will concentrate mainly on seeking for fact knowledge to solve the problem.(2) Analyzing the effect of information need on online information behavior, and comparing the effect of information need with effects of other factors from competing models. Expanding and improving the measurement of online information seeking behavior, analyzing the effect of information need on online information behavior from"the amount of seeking"dimension and"seeking process"dimension, and further comparing the effect of information need with effects of other key factors chosen from some competing perspectives or models. Such hypotheses are proposed and validated:①There exists a reverse U-shaped relationship between the amount of information need and the amount of information seeking;②The more clarifying the information need is, the less frequently the consumer will use search engine and navigation websites.③Consumers focusing on fewer websites during the earlier stage of information seeking process will expanding their access on a wider rang of websites during the later stage.④The duration of and page views in a session during the earlier stage of information seeking process are lower than which during the later stage.⑤There is a positive correlation between the amount of information seeking and the return of information seeking.⑥The width of time window of consumers'information seeking varies for different product classes. By comparison of the effects on online information seeking behavior of information need, demographic factors (age and income), product factors (product class and price), seeking return (price discount), we find information need presents more comprehensive ability to the explanation and prediction of online information seeking behavior, especially in term of the development of online information seeking behavior. In detail, the amount of information need influences significantly the amount of information seeking, but their relationship is not monotonic, seeking return comparatively is the best indicator of the amount of information seeking; As for demographic factors, neither age nor income influences significantly the amount of information seeking, although in some cases their relationship may be significant, it is still non-monotonic; For product factors, product class influences significantly the amount of information seeking, but price does not. When seeking for high-price products, the primary motivations of consumers who spending more time or accessing more websites are searching for more price discounts in nature.(3) Modeling of and analyzing on consumers'cross-site transfer behavior. Applying Zipf's law and complex network theory which are frequently adopted in researches related to internet into the modeling of consumers'cross-site transfer behavior, comparing the general law about how to allocate efforts and transfer among websites followed by consumers in a specific task context and netizens in an everyday scenario. We find cross-site behaviors conducted by consumers and netizens are common: consumers and netizens allocate their efforts among websites according to Zipf's law; the topology characteristics of their transferring networks are of great commonality, some of the characteristics (including small world, high clustering coefficient, disassortative mixing and poor invulnerability) are similar to general networks, but other characteristics are different from that of general networks, including not following power-law degree distribution, having circular and self-similar community structure. We also find that consumers'transferring between websites is highly reciprocal, and transferring mainly takes place between search engine and domain websites, especially dominantly takes place between primarily used search engine and domain websites. Consumers'transferring behavior differs on different website categories according to some indices proposed in the paper, and we can classify websites into different categories utilizing these indices. On the basis of the above finding, we put forward and implement a method which can discover automatically domain-related websites and extract macro transferring pattern among these websites. The method is operated on experiment data and real data, and in both cases it performs well and obtains practical outcome.
|
PDF Full Text Request