Search behavior analysis and evaluation of the relevance of the users of data retrieval in a web-based environment (case study)

 The objective is search behavior analysis and evaluation of relevance judgment of the users of information retrieval in the web hyperlink environment. In this regard, the pattern of the web information retrieval behavior of these users were explained and the rate of satisfaction with information retrieval system, the rate of the user trust on the results and effectiveness and ease of using the results and the rate of results relevance from user’s viewpoint was investigated. Also, the significance of the difference of the mentioned factors between the students of the field of information science and other fields was statistically assessed.

 This research is applied / fundamental which has been performed with combined method. The research population included 5784 students of master's and doctoral degree of information sciences and other different domains of sciences in Shahid Bahonar University from 2017-2018, that with purposeful sampling method, 50 students were selected. From this, 37 persons participated in the research process and the rest was omitted due to lack of cooperation in some stages of the study. This study was performed with the aid of a two-question scenario, one question relating to the academic field and one question in their specialized field. This study began through observation, searches and recording them with the aid of Camtasia Studio version 9. Then, the searches were observed and turned to text with the aid of word software. The text files were coded with MAXQDA version 10. The categories obtained from text analysis were used in explaining the users' pattern of information retrieval.
In the quantitative stage, for assessing the satisfaction rate of information retrieval system, the rate of users trusts in the results and ease of using the results and rate of relevance of the results from these user’s viewpoint, the structured interview (5 factors in the form of 15 questions in Likert scale) was performed and the data was collected and analyzed using SPSS software version 23. For evaluation of the variables, single sample t-test and for investigation of the statistical significance of the difference between two groups of information science and other fields, independent two-sample t-test was used and the average of the variables scores was compared with the mean and theoretical rate of 3. If the considered sigma score is more than the theoretical rate, it may be said that that criterion is in a desirable level. In assessing the reliability, the internal consistency coefficient above 0.7 was confirmed. Also, the validity of the interview (internal and external) was confirmed by 5 experts.

 Investigation of the search behavior analysis in completing both tasks in the qualitative section showed that among the browsers, Firefox and among the search engines, Google was more commonly used. Users completed each search in a 5-stage pattern: beginning of search, keyword codification, search navigation, retrieval and results evaluation, information extraction. In the beginning stage of more searches search engine has been used and using electronic address of the database or considered website is in the next ranks. The greatest frequency was related to simple search. In completing the first task, scienceDirect database and in the second task completion, Sivilika had the greatest use frequency. In keyword codification, the general to specific technique had the greatest frequency and after that keyword change, changing the number of words of the search term (including omission, addition and replacement), specific to general technique, simultaneous use of both general to specific and specific to general technique have been used respectively. In the search navigation, following links and menus had the greatest use. After that, following the results in the same tab, following the results in the new tab, using return key for going to the results tab and using back and forth keys are placed. In the fourth stage, often people only investigated the title of articles and congresses and reviewing the title and abstract of the papers are in the next stages of use. In the fifth stage, the subjects downloaded and saved as PDF and or their complete text could be accessed through DOI code.
The qualitative findings showed that most of the students were studying in the department of humanities and after that the group of basic sciences, technical engineering, agriculture and veterinary department. Also, the average of all investigated variables in both groups of information science and other fields has been more than average and in a desirable status. Among the variables, ease of search and achieving the results, the average scores of the information science group is in a lower level than other fields and the average score of satisfaction rate and results relevance among the students of information science is in a higher level than other fields. Also, the results showed that there is no statistically significant difference between two groups of information science group and other fields.

 Explaining the different patterns regarding the user features and the texture they are in will help the information media in responding the clients' requirement. Also, designers empower the search engines to design more interactive and user-friendly interface and instruments. Increasing the efficiency of information retrieval systems and user’s skill optimizes the search process. Intelligent retrieval systems have a considerable role in information interaction. In more and better interaction of the user with information systems, meaningful and relevant information may be obtained and this depends on perception of information systems of the users' needs and behaviors.