Pith. sign in

REVIEW 3 major objections 5 minor 51 references

An innovative data collection method to eliminate the preprocessing phase in web usage mining

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that a server-side logging API, by recording each pageview under the web server's own session identifier, eliminates the preprocessing phase of web usage mining and delivers ready-to-use structured data for pattern…

desk verdict A genuinely useful server-side logging API, but the 'eliminates preprocessing' claim is not supported; the paper needs validation before its main contribution can be taken seriously. read the letter →

arxiv 2501.04364 v1 pith:IOZB4PS5 submitted 2025-01-08 cs.IR

Ozkan Canay , Umit Kocabicak This is my paper · ORCID
classification cs.IR
keywords Webusageminingpreprocessingdatacollectionloganalysisanalyticssessionidentificationserver-sideloggingclickstream
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper's central claim is that web usage mining (WUM) does not have to start from raw server access logs. The authors propose an application-level logging API that runs on the server, captures each page request together with the web server's own session identifier, and stores the result in clean relational tables. Because user and session identification are done by the server's session mechanism rather than by offline heuristics, the paper argues that the preprocessing stage — data cleaning, user identification, sessionization, and path completion — can be skipped entirely, and pattern discovery can run directly on the stored data. A sympathetic reader would care because preprocessing is repeatedly described as the most laborious and error-prone part of WUM, and eliminating it would lower the cost and improve the reliability of clickstream analysis. The claim is supported by one day's real traffic (22,104 sessions, 161,672 pageviews) from a university's corporate web application, though the paper offers no comparison against an independent ground truth for sessions.

What carries the argument

The central object is the server-side session variable and the 'open_sessions' table that tracks live sessions. When a visitor makes the first request, the server-side programming language generates a unique alphanumeric session identifier for that client; the logging API stores session and page data under that identifier, and the record is closed when the user logs out or ignores an inactivity warning window. This mechanism is what carries the argument: it replaces the heuristic user identification (IP plus user agent) and sessionization (timeout rules such as 10 or 30 minutes) that preprocessing normally performs, and it provides the 'opn_id' foreign key that joins pageviews into ordered sessions for dwell-time and pageview-per-session calculations.

What would settle it

A targeted comparison would settle it: instrument the same site with a client-side beacon that generates an independent, per-visit identifier, run both collectors side by side on real traffic, and measure how often the server-side session boundaries disagree with the beacon's visit boundaries — especially when a user keeps a page open in one tab past the warning window, browses in multiple tabs, or hides behind a shared or rotating IP. Any mismatch shows that the session variable is not perfect ground truth, and the 'elimination of preprocessing' claim collapses unless the analyst accepts the server's sessions as definitive.

Watch

Extended reading notes

Core claim

The discovery the paper asserts is that the messy, semi-structured data problem of WUM is an artifact of using server access logs as the data source, and that a server-side logging API embedded in the application layer can replace them. The API fuses HTTP request data, network data, application data, and GeoIP data into four groups, encodes and cleans it, and writes per-session and per-page records into tables linked by a foreign key ('opn_id'). User identification is performed by the server-side session variable, so two devices behind the same dynamic IP are still recorded as separate sessions, and sessions that span midnight are not split. The paper states this yields 'one hundred percent success' in user and session identification, and that the resulting homogeneous, structured data lets a miner go directly to pattern discovery without any preprocessing sub-phase.

Load-bearing premise

The method assumes the web server's session variable, together with an inactivity timeout warning, correctly identifies each real human visitor and each visit, so the recorded session boundaries are true ground truth rather than another heuristic.

Editorial extensions

If this is right

  • Organizations that adopt this method can feed pattern discovery algorithms directly from the relational tables, skipping data filtering, user identification, sessionization, and path completion.
  • The data doubles as a real-time monitoring source (open_sessions shows active visitors) and as a foundation for web analytics, machine learning, and deep learning inputs without ETL-like cleaning.
  • Because data collection is server-side, there is no second JavaScript request per page transition, so the server is less busy and the data captures search-engine bots more accurately than client-side tagging.
  • Dynamic IP addresses and sessions crossing midnight no longer confuse user and session attribution, addressing known weaknesses of IP-based log heuristics.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The claim that preprocessing is wholly eliminated depends on the session variable being a perfect ground truth; a natural stress test is to compare session boundaries from this method against a client-side beacon with an independent visit ID, which the paper does not report.
  • A likely boundary condition: a user who leaves a page open in a tab past the warning window and returns triggers a new session, so 'one visit' can still be split despite the 100% success claim; the paper does not quantify how often this happens.
  • The paper's own conclusion admits the operational OLTP database is write-heavy and unsuitable for mining queries, implying that in practice an extraction or warehouse step reappears — the preprocessing eliminated at collection time may re-enter as an ETL step for long-term analysis.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes an application-level, server-side logging API as an alternative to web server access logs for web usage mining (WUM). The method is implemented as part of a three-tier web application framework; for each page request it collects HTTP, network, application, and GeoIP data, stores them in relational tables (session and pageview logs), and manages sessions via server-side session variables with an inactivity-based 'warning window' timeout. The authors claim that this approach eliminates the WUM preprocessing stage (data cleaning, filtering, user identification, session identification, path completion, and formatting) and that user and session identification is achieved with 'one hundred percent success' (Section 3.2). To demonstrate the method, they present descriptive analytics from 24 hours of real data from a university web application, comprising 22,104 sessions and 161,672 pageviews, along with breakdowns by user type, gender, device, browser, IP, and search engine.

Significance. If the central claim were established, the method would be a practically valuable contribution: it would remove a notoriously labor-intensive step from WUM, give organizations full ownership of structured usage data, and avoid the second-request overhead of client-side tracking. The paper has real strengths: the system is implemented and deployed on a production university web application, the data model is concrete, a sample tuple is shown, and the data collection process is described with a sequence diagram and a complexity statement. These elements make the engineering narrative credible. However, the scientific claim that preprocessing is 'eliminated' is not supported by the evidence presented. The evaluation is purely descriptive and self-referential; it shows that analytics can be computed from the collected data, not that session boundaries or user identities are correct, nor that mining outcomes are improved relative to a baseline. The one-hundred-percent identification claim is contradicted by the method's own dependence on cookie-keyed sessions and by the unavoidable loss of cache/back requests.

major comments (3)
  1. [Section 3 (page 7) and Section 4 (page 10)] The paper's central claim that the method 'eliminates the need for the preprocessing stage' and 'presents healthy, clean, and ready-to-use data' for pattern discovery is a comparative/quantitative claim, but Section 4 provides no such comparison. There is no baseline against traditional server-log preprocessing, no ground-truth session labels, no accuracy metric for user or session identification, and no downstream mining task (association, clustering, classification) demonstrating that the data are directly usable. The reported aggregates (Tables 4-6, Figures 5-8) show only that descriptive statistics can be extracted from the collected data, which is not sufficient to establish elimination of preprocessing.
  2. [Section 3.2 (page 9)] The justification for the 'one hundred percent success' of user and session identification—'at least one data in the table row changes on each page requested'—is a non sequitur. The session variable is maintained by the server-side language via cookies, so it inherits standard identity ambiguities: cookie deletion or browser switching creates multiple sessions for one user, while shared devices or browser profiles merge distinct users. More importantly, pages served from the browser cache or bfcache during 'Back' navigation never reach the server and therefore never invoke the API; the recorded path is incomplete, and path completion is not eliminated, only omitted. The inactivity 'warning window' described in Section 3 is a heuristic session boundary, not a ground truth. No experiment in Section 4 compares the method's session output with a known session log, so the 100% claim is empirically unsupported and, for cache/back cases, internally inconsistent with the proposed mechanism.
  3. [Section 5 (page 14)] The conclusion states that 'completely clean and hassle-free' data are available for mining, but the same section concedes that the operational write-heavy database is 'not suitable for performing queries on it and executing mining operations' and that an ETL process to a data warehouse is needed before mining. This is a significant qualification of the 'ready-to-use' claim and should be reflected in the abstract and in Section 3, where the elimination of preprocessing is announced. The experimental section does not present any timing, resource-usage, or effort comparison with a conventional log-based pipeline, so the asserted acceleration of the WUM process (Section 1, contribution 4) also lacks quantitative support.
minor comments (5)
  1. [Section 3.3, Eq. (2)] The dwell-time formula D_s = sum_{p=1}^{m} (R_{p+1} - R_p) is undefined for the last page in the session (R_{m+1} does not exist), and it also does not account for the final page's dwell time or for the timeout interval after the last request. Since Table 5 reports total session durations, this undefined term affects the reported numbers and should be clarified.
  2. [Section 4.1, Table 4] The classification of 'Guests' versus 'Users' would benefit from a statement of how the two groups are identified in the data; if it is based on the log_uid field, the claim of 'one hundred percent success' for user identification should be explicitly qualified to exclude situations where no login is present.
  3. [Abstract and Section 1] The statement that the method 'works faster than client-side methods and makes the server less busy' is not substantiated by any benchmark or load measurement; the O(N) statement in Section 3.1 is about scaling, not about comparative speed, and the claim should be toned down or supported with data.
  4. [Section 2.6 (page 5)] In the formal definition of a log entry, the symbol 'c' is used both for the HTTP status code and for the size of transferred bytes, which is confusing; a distinct symbol for one of the two quantities would improve readability.
  5. [Global] Several typos and minor errors are present: the title on the first page reads 'W eb Usage Mining'; 'relation algebra' should be 'relational algebra' in Section 3.3; and references [7] and [24] appear to be duplicated. The authors should also cite the prior conference paper [52] explicitly when describing what is new in this journal version.

Circularity Check

1 steps flagged · score 6.0 of 10

The '100% success' user/session identification claim is tautological: sessions are defined as the server's own session variable, so identification cannot fail by construction; this definitional step is load-bearing for the paper's core 'preprocessing eliminated' claim.

  1. self definitional [Section 3.3 ('Data cleaning, processing, and storage'), relying on the session-variable definition in Section 3.2 ('Data model')]
    "This method has been preferred since at least one data in the table row changes on each page requested. Thus, the user and session identification would be ensured with one hundred percent success."

    The 100% success claim is not derived from any external ground truth; it is true by construction because the paper defines a session as the application server's automatically generated session variable. Section 3.2 states 'a new alphanumeric session variable is generated automatically for the client by the server-side programming language,' and Section 3 states 'sessions are created by the web server.' The identifier used to demonstrate 'user and session identification' is therefore the method's own output. No comparison against known sessions, no reconstruction check, and no accuracy metric is reported; instead the WUM sessionization problem is dissolved by redefining sessions as whatever the server session variable tracks.

full rationale

The only defensible circularity finding is the paper's 100% user/session-identification claim in Section 3.3. The paper treats the application server's session variable as the definition of a session, then reports that identification is 100% successful; this is a tautology, since success is measured against the method's own identifier. This step is load-bearing: the 'eliminates the need for the preprocessing stage' claim in Section 3 depends on user/session identification having been solved, and the paper offers no independent validation (no known-session benchmark, no cache/back-navigation reconstruction check, no cookie-reset analysis). Because of this construction, the claim is partially circular. However, the rest of the contribution is not circular: no parameters are fitted to data and then re-predicted; the descriptive statistics in Section 4 merely illustrate the collected data; the self-citation [52] is a background citation about content providers' interest in visitor behavior, not a load-bearing uniqueness theorem. The server-side logging architecture itself has independent engineering content, and the preprocessing work is, by the paper's own description, relocated into the logging API rather than hidden. Thus the circularity is confined to the identification-success claim and the preprocessing-elimination conclusion that leans on it, warranting a mid-scale score rather than a higher one.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The paper introduces no fitted model parameters; its only hidden knob is the session inactivity timeout, which is not disclosed. The load-bearing assumptions are unvalidated claims about server-side session identifiers being ground truth, full capture of page requests by the framework, and linear dwell-time computation, none of which are checked against a baseline.

free parameters (1)
  • session_inactivity_timeout
    Session-ending inactivity threshold is mentioned but never quantified; controls all session-level statistics.
assumptions (3)
  • domain assumption Server-side session identifiers give a correct and complete map of users to visits.
    Underpins '100 percent success' in Section 3.2; unvalidated.
  • domain assumption The logging API is invoked for every user page request because it is part of the framework head.
    Section 3.1 states 'the API is also called for every page request made'; static, cached, or cross-domain requests may be missed.
  • domain assumption Page dwell time is the difference between consecutive request times in the same session.
    Eq. (2) in Section 3.3 assumes linear single-tab browsing; multi-tab and cached back-navigation break the assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An innovative data collection method to eliminate the preprocessing phase in web usage mining." pith.science (2026). https://pith.science/paper/IOZB4PS5

@misc{pith2026250104364,
  author       = {Pith},
  title        = {Pith review of: An innovative data collection method to eliminate the preprocessing phase in web usage mining},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IOZB4PS5}},
  note         = {Machine review of arXiv:2501.04364}
}
read the original abstract

The underlying data source for web usage mining (WUM) is commonly thought to be server logs. However, access log files ensure quite limited data about the clients. Identifying sessions from this messy data takes a considerable effort, and operations performed for this purpose do not always yield excellent results. Also, this data cannot be used for web analytics efficiently. This study proposes an innovative method for user tracking, session management, and collecting web usage data. The method is mainly based on a new approach for using collected data for web analytics extraction as the data source in web usage mining. An application-based API has been developed with a different strategy from conventional client-side methods to obtain and process log data. The log data has been successfully gathered by integrating the technique into an enterprise web application. The results reveal that the homogeneous structured data collected and stored with this method is more convenient to browse, filter, and process than web server logs. This data stored on a relational database can be used effortlessly as a reliable data source for high-performance web usage mining activity, real-time web analytics, or a functional recommendation system.

Figures

Figures reproduced from arXiv: 2501.04364 by the authors.

Figure 1
Figure 1. Proposed data collection method and improvement of [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Three-tier organization of the proposed method [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Sequence diagram representation of the way the log API [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: The entity-relationship diagram of the physical data model [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: Top 20 users’ numbers of pageviews and sessions by the most pageviews in all sessions [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: Hourly user numbers of pageviews by user types [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: Technical analysis of client-oriented device and browser [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Frequency of search engines and search keywords [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 45 canonical work pages

  1. [1]

    V.Jain, K.Kashyap, Anefficientalgorithmforweblogdatapre- processing, in: Machine Vision and Augmented Intelligence— Theory and Applications, Springer Singapore, Singapore, 2021, p. 505–514. doi: 10.1007/978-981-16-5078-9_41

    work page doi:10.1007/978-981-16-5078-9_41 2021

  2. [2]

    Abdalla, T

    A. Abdalla, T. Ahmed, M. Seliaman, Web usage mining and the challenge of big data: A review of emerging tools and techniques, in: I.R.M. Association (Ed.), Big Data: Concepts, Methodologies, Tools, and Applications, vol. 6, IGI Global, 2016, Ch. 42, p. 899–928. doi: 10.4018/978-1-4666-9840- 6.ch042

    work page doi:10.4018/978-1-4666-9840- 2016

  3. [4]

    Čegan, P

    L. Čegan, P. Filip, Webalyt: Open web analytics platform, in: 2017 27th International Conference Radioelektronika, IEEE, 2017, p. 1–5. doi: 10.1109/RADIOELEK.2017.7937605

    work page arXiv 2017

  4. [5]

    Y. Tao, S. Guo, C. Shi, D. Chu, User behavior analysis by cross-domain log data fusion, IEEE Access 8 (2019) 400–406. doi: 10.1109/ACCESS.2019.2961769

    work page arXiv 2019

  5. [6]

    Ehikioya, S

    S.A. Ehikioya, S. Lu, A path analysis model for effective e- commerce transactions, African Journal of Computing and ICT 12 (2) (2019) 55–71

    work page 2019

  6. [7]

    Roy, G.A

    R. Roy, G.A. Rao, Survey on pre-processing web log files in web usage mining, International Journal of Advanced Science and Technology 29 (3 Special Issue) (2020) 682–691

    work page 2020

  7. [8]

    Ibrahim, A.J

    K.K. Ibrahim, A.J. Obaid, Web mining techniques and tech- nologies: A landscape view, Journal of Physics: Confer- ence Series 1879 (3) (2021) 032125. doi: 10.1088/1742- 6596/1879/3/032125

    doi:10.1088/1742- 2021

  8. [9]

    Srivastava, A.K

    M. Srivastava, A.K. Srivastava, R. Garg, P. Mishra, Perfor- mance evaluation of the mapreduce-based parallel data pre- processing algorithm in web usage mining with robot de- tection approaches, IETE Technical Review (2021) 1–15doi: 10.1080/02564602.2021.1918584

    arXiv 2021

Show all 51 references
  1. [10]

    Bayir, I.H

    M.A. Bayir, I.H. Toroslu, Maximal paths recipe for con- structing web user sessions, World Wide Web (2022) 1–31doi: 10.1007/s11280-022-01024-3

    2022 doi

  2. [11]

    M. Munk, L. Benko, Using entropy in web usage data prepro- cessing, Entropy 20 (1) (2018) 67. doi: 10.3390/e20010067

    2018 doi

  3. [13]

    Knight-Davis, Using awstats to analyze logs from ezproxy and from the public opac logs, in: Spring Forum: Collection Management and Technical Services Committees, 2017, p

    S. Knight-Davis, Using awstats to analyze logs from ezproxy and from the public opac logs, in: Spring Forum: Collection Management and Technical Services Committees, 2017, p. 228

    2017

  4. [14]

    Gamalielsson, B

    J. Gamalielsson, B. Lundell, S. Butler, C. Brax, T. Persson, A. Mattsson, T. Gustavsson, J. Feist, E. Lönroth, Towards open government through open source software for web ana- lytics: The case of matomo, JeDEM-eJournal of eDemocracy and Open Government 13 (2) (2021) 133–153. ...

    2021 doi

  5. [15]

    Aartsen, O.F

    B. Aartsen, O.F. El-Gayar, C. Noteboom, A systematic review of web usage mining techniques and future research options, in: MWAIS 2020 Proceedings, vol. 25, MWAIS, 2020, pp. 1–6

    2020

  6. [16]

    Milosevic, D

    B. Milosevic, D. Regodic, V. Saso, Big data management pro- cesses in business intelligence systems, in: Economic and Social Development: Book of Proceedings, Varazdin Development and Entrepreneurship Agency (VADEA), 2021, pp. 182–192

    2021

  7. [17]

    Zheng, S

    G. Zheng, S. Peltsverger, Web analytics overview, in: Encyclo- pedia of Information Science and Technology, Third Edition, IGI Global, 2015, p. 7674–7683. doi: 10.4018/978-1-4666-5888- 2.ch756

    2015 doi

  8. [18]

    Srivastava, R

    M. Srivastava, R. Garg, P.K. Mishra, Analysis of data extrac- tion and data cleaning in web usage mining, in: Proceedings of the 2015 International Conference on Advanced Research in Computer Science Engineering & Technology (ICARCSET 2015), Association for Computing Machinery...

    2015

  9. [19]

    Mishra, S

    S. Mishra, S. Srivastava, Web development frameworks and its performance analysis—a review, Smart Computing (2021) 337–343doi: 10.1201/9781003167488-39

    2021 doi

  10. [20]

    Clifton, Advanced web metrics with google analytics, in: Ad- vanced Web Metrics with Google Analytics, 3rd Edition, John Wiley & Sons: Indianapolis, IN, USA, 2012, pp

    B. Clifton, Advanced web metrics with google analytics, in: Ad- vanced Web Metrics with Google Analytics, 3rd Edition, John Wiley & Sons: Indianapolis, IN, USA, 2012, pp. 3–5

    2012

  11. [21]

    Onder, A

    I. Onder, A. Berbekova, Web analytics: more than website per- formance evaluation?, International Journal of Tourism Cities- doi: 10.1108/IJTC-03-2021-0039

    2021 doi

  12. [22]

    Reddy, A review of data warehouses multidimensional model and data mining, Information Technology in Industry 9 (3) (2021) 310–320

    G. Reddy, A review of data warehouses multidimensional model and data mining, Information Technology in Industry 9 (3) (2021) 310–320

    2021

  13. [24]

    Shah, H.B

    P. Shah, H.B. Pandit, A review: Web content mining tech- niques, Data Engineering for Smart Systems (2022) 159–172doi: 10.1007/978-981-16-2641-8_15

    2022 doi

  14. [25]

    Tyagi, S.K

    N. Tyagi, S.K. Gupta, Web structure mining algorithms: A survey, in: Big Data Analytics, Springer, 2018, pp. 305–317. doi: 10.1007/978-981-10-6620-7_30

    2018 doi

  15. [26]

    Lim, L.Y

    Z.Y. Lim, L.Y. Ong, M.C. Leow, A review on clustering tech- niques: Creating better user experience for online roadshow, 14 Future Internet 13 (9) (2021) 233. doi: 10.3390/fi13090233

    2021 doi

  16. [27]

    R. Das, I. Turkoglu, Creating meaningful data from web logs for improving the impressiveness of a website by using path anal- ysis method, Expert Systems with Applications 36 (3) (2009) 6635–6644. doi: 10.1016/j.eswa.2008.08.067

    2009 doi

  17. [28]

    Manchanda, N

    M. Manchanda, N. Gupta, Web usage mining: Dynamic methodology to preprocessing web logs, HELIX 8 (5) (2018) 3810–3815. doi: 10.29042/2018-3810-3815

    2018 doi

  18. [29]

    Jokar, A.R

    N. Jokar, A.R. Honarvar, S. Aghamirzadeh, K. Esfandiari, Web mining and web usage mining techniques, Bulletin de la Société desSciencesdeLiège85(1)(2016)321–328. doi: 10.25518/0037- 9565.5371

    2016

  19. [30]

    Kumar, R

    S. Kumar, R. Kumar, A study on different aspects of web min- ing and research issues, in: IOP Conference Series: Materials Science and Engineering, vol. 1022, IOP Publishing, 2021, pp. 012–018. doi: 10.1088/1757-899X/1022/1/012018

    2021 doi

  20. [31]

    Kewen, Analysis of preprocessing methods for web usage data, in: Proceedings of 2012 International Conference on Mea- surement, Information and Control, vol

    L. Kewen, Analysis of preprocessing methods for web usage data, in: Proceedings of 2012 International Conference on Mea- surement, Information and Control, vol. 1, IEEE, 2012, p. 383–386. doi: 10.1109/MIC.2012.6273276

    2012

  21. [32]

    Mobasher, Web mining overview, in: J

    B. Mobasher, Web mining overview, in: J. Wang (Ed.), Ency- clopedia of Data Warehousing and Mining, Second Edition, 2nd Edition, vol. 3, IGI Global, 2009, Ch. 319, pp. 2085–2089. doi: 10.4018/978-1-60566-010-3.ch319

    2009 doi

  22. [33]

    Slanzi, G

    G. Slanzi, G. Pizarro, J.D. Velásquez, Biometric informa- tion fusion for web user navigation and preferences analy- sis: An overview, Information Fusion 38 (2017) 12–21. doi: 10.1016/j.inffus.2017.02.006

    2017 doi

  23. [34]

    Nandal, et al., A systematic review on data preprocess- ing and pattern discovery of web usage mining, International Journal of Advanced Research in Computer Science 9 (2)

    R. Nandal, et al., A systematic review on data preprocess- ing and pattern discovery of web usage mining, International Journal of Advanced Research in Computer Science 9 (2). doi: 10.26483/ijarcs.v9i2.5763

    doi

  24. [35]

    Fatima, H

    B. Fatima, H. Ramzan, S. Asghar, Session identification tech- niquesusedinwebusagemining: asystematicmappingofschol- arlyliterature, OnlineInformationReviewdoi: 10.1108/OIR-08- 2015-0274

    2015 doi

  25. [36]

    Joachims, L

    T. Joachims, L. Granka, B. Pan, H. Hembrooke, G. Gay, Ac- curately interpreting clickthrough data as implicit feedback, in: ACM SIGIR Forum, vol. 51, ACM, New York, NY, USA, 2017, p. 4–11. doi: 10.1145/1076034.1076063

    2017

  26. [37]

    García, J

    S. García, J. Luengo, F. Herrera, Data preprocessing in data mining, vol. 72, Springer, 2015. doi: 10.1007/978-3-319-10247- 4

    2015 doi

  27. [38]

    N. Kaur, H. Aggarwal, A novel semantically-time-referrer based approach of web usage mining for improved session- ization in pre-processing of web log, International journal of advanced computer science and applications 8 (1). doi: 10.14569/IJACSA.2017.080122

  28. [39]

    Srivastava, R

    M. Srivastava, R. Garg, P. Mishra, A mapreduce-based user identification algorithm in web usage mining, In- ternational Journal of Information Technology and Web Engineering (IJITWE) 13 (2) (2018) 11–23. doi: 10.4018/IJITWE.2018040102

    2018 doi

  29. [40]

    Fatima, H

    B. Fatima, H. Ramzan, S. Asghar, Session identification tech- niques used in web usage mining: A systematic mapping of scholarly literature, Online Information Review 40 (7) (2016) 1033–1053. doi: 10.1108/OIR-08-2015-0274

    2016 doi

  30. [41]

    Mughal, Data mining: Web data mining techniques, tools and algorithms: An overview, International Journal of Advanced Computer Science and Applications 9 (6)

    M.J.H. Mughal, Data mining: Web data mining techniques, tools and algorithms: An overview, International Journal of Advanced Computer Science and Applications 9 (6). doi: 10.14569/IJACSA.2018.090630

    2018

  31. [42]

    Clifton, Advanced web metrics with Google Analytics, John Wiley & Sons, 2012

    B. Clifton, Advanced web metrics with Google Analytics, John Wiley & Sons, 2012

    2012

  32. [43]

    Varnagar, N.N

    C.R. Varnagar, N.N. Madhak, T.M. Kodinariya, J.N. Rathod, Web usage mining: A review on process, methods and tech- niques, in: 2013 International Conference on Information Com- munication and Embedded Systems (ICICES), IEEE, 2013, pp. 40–46. doi: 10.1109/ICICES.2013.6508399

    2013

  33. [44]

    Gholamian, P

    S. Gholamian, P. Ward, A comprehensive survey of log- ging in software: From logging statements automation to log mining and analysis, arXiv preprint arXiv:2110.12489doi: 10.48550/arXiv.2110.12489

    doi arXiv

  34. [45]

    Srivastava, A

    M. Srivastava, A. Srivastava, R. Garg, Data preprocessing tech- niques in web usage mining: A literature review, in: Pro- ceedings of International Conference on Sustainable Computing in Science, Technology and Management (SUSCOM), Amity University Rajasthan, Jaipur-India, 201...

    2019 doi

  35. [46]

    Paredes, J

    R. Paredes, J. Bolanio, Analyzing logs from proxy server and captive portal using k-means clustering algorithm, Middle East Journal of Applied Science & Technology 3 (4) (2020) 10–31. doi: 10.4018/IJCBPL.2020010101

    2020 doi

  36. [47]

    Deshpande, S

    D. Deshpande, S. Deshpande, V. Thakare, Web user identifi- cation: Analysis of heuristic solutions, in: 2018 Second Inter- national Conference on Intelligent Computing and Control Sys- tems (ICICCS), IEEE, 2018, p. 1790–1795. doi: 10.1109/IC- CONS.2018.8662893

    2018

  37. [48]

    Sukumar, L

    P. Sukumar, L. Robert, S. Yuvaraj, Review on modern data preprocessing techniques in web usage mining (wum), in: 2016 International Conference on Computation System and Infor- mation Technology for Sustainable Solutions (CSITSS), IEEE, 2016, p. 64–69. doi: 10.1109/CSITSS.2016.7779441

    2016

  38. [49]

    Kundu, L

    S. Kundu, L. Garg, Web log analyzer tools: A comparative study to analyze user behavior, in: 2017 7th International Conference on Cloud Computing, Data Science & Engineering- Confluence, IEEE, 2017, p. 17–24. doi: 10.1109/CONFLU- ENCE.2017.7943117

    2017

  39. [50]

    P. Svec, L. Benko, M. Kadlecik, J. Kratochvil, M. Munk, Web usage mining: Data pre-processing impact on found knowledge in predictive modelling, Procedia Computer Science 171 (2020) 168–178. doi: 10.1016/j.procs.2020.04.018

    2020 doi

  40. [51]

    Quintel, R

    D. Quintel, R. Wilson, Analytics and privacy, Informa- tion Technology and Libraries 39 (3) (2020) 1–11. doi: 10.6017/ital.v39i3.12219

    2020 doi

  41. [52]

    Canay, U

    O. Canay, U. Kocabicak, A new data collection model for in- formation extraction from web click logs, in: International Ar- tificial Intelligence and Data Processing Symposium (IDAP), Inonu University, Malatya, Turkey, 2016, p. 489–492

    2016

  42. [53]

    Canay, S

    O. Canay, S. Meric, H. Evirgen, M. Varan, Realization of cam- pus automation web information system in context of service unity architecture, in: International Symposium on Comput- ing in Science & Engineering (ISCSE), Izmir, Turkey, 2011, p. 173–179

    2011

  43. [54]

    Malarvizhi, B

    S. Malarvizhi, B. Sathiyabhama, Frequent pagesets from web log by enhanced weighted association rule mining, Cluster Com- puting 19 (1) (2016) 269–277. doi: 10.1007/s10586-015-0507-z. 15

    2016 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.