CoChat
Sign Up
Cited by 27Remote-Sensing Image ClassificationSpectroscopy and Chemometric AnalysesAdvanced Neural Network Applications

Development of online classification system for construction waste based on industrial camera and hyperspectral camera

Open access
Xiao, Wen, Yang, Jianhong, Fang, Huaiying, Zhuang, Jiangteng, Ku, Yuedong·2019·PLOS ONE
DOI 10.1371/journal.pone.0208706PMC PMC6334961
Open this paper in CoChat — free accountOpen PDF

Abstract

Construction waste is a serious problem that should be addressed to protect environment and save resources, some of which have a high recovery value. To efficiently recover construction waste, an online classification system is developed using an industrial near-infrared hyperspectral camera. This system uses the industrial camera to capture a region of interest and a hyperspectral camera to obtain the spectral information about objects corresponding to the region of interest. The spectral information is then used to build classification models based on extreme learning machine and resemblance discriminant analysis. To further improve this system, an online particle swarm optimization extreme learning machine is developed. The results indicate that if a near-infrared hyperspectral camera is used in conjunction with an industrial camera, construction waste can be efficiently classified. Therefore, extreme learning machine and resemblance discriminant analysis can be used to classify construction waste. Particle swarm optimization can be used to further enhance the proposed system.

AI snapshot

Written from the paper’s full text.

TL;DR

The paper develops an online construction-waste classification system combining a 2D industrial camera for ROI detection and a NIR hyperspectral camera (900–1700 nm) for spectral classification. It reports that RDA achieved 100% accuracy with PCA or SPA features, while ELM reached 96.667–98% and PCA–PSO–OL–ELM reached 98.667% at much higher training cost.

Key findings

  • RDA with either PCA or SPA feature bands classified all test pieces correctly: 30/30 for each of foam, plastic, brick, concrete, and wood, giving 100% overall accuracy with a 28 ms training time.
  • ELM gave 96.667% overall accuracy with SPA features and 98% with PCA features; SPA errors mainly occurred in brick and concrete, while PCA improved brick and concrete but slightly decreased foam.
  • PCA–PSO–OL–ELM improved overall accuracy to 98.667% (29/30 foam, 30/30 plastic, 30/30 brick, 29/30 concrete, 30/30 wood) but required 1550.83 s training time; average similarity increased for foam, brick, concrete, and wood, while plastic decreased from 61.99% to 60.00%.
  • Feature selection used PCA, whose three principal components explained 98.16% of information (PC1 77.69%, PC2 14.86%, PC3 5.61%), yielding bands at 1215, 1296, 1391, and 1433 nm; SPA yielded bands at 1084, 1215, 1345, and 1511 nm.

Methods

  • The system runs offline model building and online classification: an industrial camera captures a 2D image, a 5 × 5 pixel ROI is taken around the object centroid, and a NIR hyperspectral camera collects spectral data in that ROI; belt speed was 115 mm/s, industrial camera frame rate 10 fps, and hyperspectral camera frame rate 120 fps.
  • The paper states that 80 pieces each of plastics, bricks, concretes, and woods were obtained, with 50 used for training and 30 for testing; the results and tables instead report five classes—foam, plastic, brick, concrete, and wood—with 50 training and 30 testing pieces per class.
  • Preprocessing used white/black correction and Savitzky–Golay 7-point smoothing; the first and last 12 bands were excluded due to noise. PCA and SPA extracted feature bands, and classification used RDA, ELM, and PCA–PSO–OL–ELM, with PSO optimizing ELM weights/biases and online learning adding new data at a 70% similarity threshold over 500 iterations.

Limitations

  • Stated: the testing data size was small and did not reveal the overall accuracy; PSO was time-consuming; RDA is rigid and may perform poorly with insufficiently representative samples or complicated nonlinear problems; ELM needs sufficient training data; and the distance/angle weight was fixed at 1:1 and may need adjustment in other applications.
  • Evident: the sample description says four materials (plastics, bricks, concretes, woods) with 80 pieces each, but the results and tables include five classes including foam and report 50 training/30 testing pieces per class; this inconsistency is not reconciled in the text.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0208706

Cites

13 of 13 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

2 of 17 cited sources checked · 1 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    Foam, plastic, and wood are organic, with obvious absorption peaks expressed as trough characteristics on the average reflectance curve.

    “Foam, plastic, wood are organic, with obvious absorption peaks, expressed as trough characteristics on the curve, while brick and concrete are inorganic, without obvious absorption peak characteristics.”
    Quote in the paperSubstantial claim
  • Findingown data

    The five types of materials can be clearly distinguished around 1200 and 1400 nm.

    “The five types of materials can be clearly distinguished around 1200 and 1400 nm.”
    Quote in the paperSubstantial claim
  • Findingreasoning

    The ELM algorithm does not yield enough accurate results.

    “These techniques enhanced the efficiency of ELM; however, the ELM algorithm does not yield enough accurate results.”
    Quote in the paperSubstantial claim
  • Findingown data

    As shown in Table 1, 117 ms are required in 250 training datasets.

    “As shown in Table 1, 117 ms are required in 250 training datasets.”
    Quote in the paperNot a substantial claim
  • Findingown data

    The training time will increase with the increasing amount of training data.

    “The training time will increase with the increasing amount of training data.”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim. 16
  • May combine two claims. 24
  • The quote only partly states this claim. 1

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The most recent ways involving classification using hyperspectral data include extreme learning machine (ELM), partial least squares discriminant analysis (PLS-DA), and various deep learning algorithms.

    “The most recent ways involving classification using hyperspectral data include extreme learning machine (ELM) [7–9], partial least squares discriminant analysis (PLS-DA) [10,11], and various deep learning algorithms [12,13].”
    Quote in the paperNot a substantial claim2 of 7 sources checked
    • cites [7] Guo C. Extreme learning machine and its application in classification. Doctoral dissertation, Zhengzhou University. 2013.

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [8] DB Heras, 2014, Exploring elm-based spatial–spectral classification of hyperspectral images, Int J Remote Sens source
    • cites [9] F Argüello, 2015, Elm-based spectral–spatial classification of hyperspectral images using extended morphological profiles and composite feature mappings, Int J Remote Sens source
    • cites [10] E Ivorra, 2013, Detection of expired vacuum-packed smoked salmon based on pls-da method using hyperspectral images, J Food Eng source
    • cites [11] I Marquetti, 2016, Partial least square with discriminant analysis and near infrared spectroscopy for evaluation of geographic and genotypic origin of arabica coffee, Comput Electron Agric source
    • cites [12] JH Le, 2015, International Symposium on Visual Computing

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [13] F Palsson, 2047, Ulfarsson MO Multispectral and hyperspectral image fusion using a 3-d-convolutional neural network, IEEE Geosci. Remote Sens. Lett source

      The Introduction (p. 1) describes a deep-learning method for fusing multispectral and hyperspectral images, so the paper supports the deep-learning portion of the citation. It identifies classification as a possible use of hyperspectral information but does not present its method as hyperspectral classification; it does not support that qualification. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given

    The edge points of various materials are similar in some degree.

    “The edge points of various materials are similar in some degree.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    The five species are distinguished according to their spectral data, but some species could not be classified because they resembled wood, foam, brick, and some species such as plastic and concrete.

    “These five species are distinguished according to their spectral data. However, some species could not be classified because they resembled wood, foam, brick, and some species such as plastic and concrete.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    The AS of foam, brick, and concrete changed a lot, while the AS of plastic and wood changed less.

    “The AS of foam, brick, and concrete has changed a lot, while the AS of plastic and wood has changed less.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    The testing time was relatively shorter compared to the time consumed during training when PSO was working.

    “The testing time was relatively shorter compared to the time consumed during training when the PSO was working.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    ELM in combination with PCA can yield 98% accurate results, but this result is lower than RDA.

    “Although ELM in combination with PCA can yield 98% accurate results, this result is lower than RDA.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    With PCA, the correct rate of bricks and concrete improved, but the correct rate of foam decreased slightly.

    “Although the correct rate of bricks and concrete has been improved, at the same time, the correct rate of foam has decreased slightly.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    The process of PSO-OL-ELM was time-consuming.

    “However, the process was time-consuming.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Findingown data

    Except for plastics, the remaining average similarity (AS) of species increased with PSO-OL-ELM compared to ELM.

    “Except for plastics, the remaining AS of species have increased.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The three principal components can provide 98.16% of the information about the original data, with PC1 (77.69%), PC2 (14.86%), and PC3 (5.61%).

    “The three principal components can provide 98.16% of the information about the original data. Specifically, PC1 (77.69%), PC2 (14.86%), and PC3 (5.61%), which can show original features in some degree.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    The accuracy of plastic and wood are both 100%, whose capability of optimization are lower.

    “However the accuracy of plastic and wood are both 100%, whose capability of optimization are lower.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The overall accuracy of ELM with PCA is 98%, slightly higher than with SPA.

    “The overall accuracy with PCA is 98%, slightly higher than the SPA.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Study design and conductreasoning

    When new data has a similarity threshold of 70%, it will be incorporated as a part of the training data because it enhances the accuracy of the testing procedure.

    “When new data has a similarity threshold of 70%, it will be incorporated as a part of the training data because it enhances the accuracy of the testing procedure.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The size of the training set affects the accuracy.

  • May combine two claims.

    Premisecitation

    The principle component analysis (PCA) enables extracting spectral feature wavebands and yields good result.

    “The principle component analysis (PCA) [3] enables extracting spectral feature wavebands and yields good result.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [3] S Serranti, 2012, Classification of polyolefins from building and construction waste using nir hyperspectral imaging system, Resour Conserv Recycl source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    The successive projections algorithm (SPA) separates original matrix to several small matrices and obtains feature wavebands.

    “The successive projections algorithm (SPA) [6] separates original matrix to several small matrices.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [6] Q Ni, 2016, Red rosewood wood species identification based on hyperspectral imaging technology, Journal of Zhejiang Agricultural and Forestry University

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationown data

    Extreme learning machine and resemblance discriminant analysis can be used to classify construction waste.

    “Therefore, extreme learning machine and resemblance discriminant analysis can be used to classify construction waste.”
    • Also stated in: § Conclusions · boosted
      “We then successfully used RDA and ELM to discriminate construction waste.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Some studies have used near-infrared hyperspectral cameras to classify different objects.

    “Some studies have used near-infrared hyperspectral camera to classify different objects [2,3].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [2] EV Dmitriev, 2014, Classification of the forest cover of tver oblast using hyperspectral airborne imagery, Zv Atmos Ocean Phy source
    • cites [3] S Serranti, 2012, Classification of polyolefins from building and construction waste using nir hyperspectral imaging system, Resour Conserv Recycl source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Study design and conductnone given

    Using the Savitsky-Golay (SG) algorithm to denoise the spectral curve can achieve good results and it is easier to implement.

    “Using the Savitsky–Golay (SG) algorithm to denoise the spectral curve can achieve good results and it is easier to implement.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    The original data required correction by white data and black data to ensure a standard reflectance and minimize variations influenced by light source differences.

    “The original data required correction by white data and black data to ensure a standard reflectance and minimize variations influenced by light source differences.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Several other studies have focused on further improving the classification accuracy of hyperspectral data.

    “Several other studies have focused on further improving the classification accuracy [14–16].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [14] B Demir, 2010, Empirical mode decomposition based decision fusion for higher hyperspectral image classification accuracy, Geoscience and Remote Sensing Symposium source
    • cites [15] RYM Nakamura, 2014, Nature-inspired framework for hyperspectral band selection, IEEE Trans. Geosci. Remote Sens source
    • cites [16] J Wang, 2013

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The ELM is a simple, single hidden-layer forward neural network.

    “The ELM is a simple, single hidden-layer forward neural network that was developed by professor Hunang [18].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [18] A Savitzky, 1964, Smoothing and Differentiation of Data by Simplified Least Squares Procedures, Anal Chem source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The ELM can be quickly trained without compromising the accuracy of the results.

    “This algorithm can be quickly trained without compromising the accuracy of the results.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [18] A Savitzky, 1964, Smoothing and Differentiation of Data by Simplified Least Squares Procedures, Anal Chem source

  • Classified as probably not a substantial claim.

    Premisecitation

    The crests and troughs of the curve of load coefficient can express the difference of original curve.

    “The crests and troughs of the curve of load coefficient can express the difference of original curve [3].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [3] S Serranti, 2012, Classification of polyolefins from building and construction waste using nir hyperspectral imaging system, Resour Conserv Recycl source

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    This system is dependable to some extent.

    “Therefore, this system is dependable to some extent.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationnone given

    When representative samples are insufficiently representative, the correct rate may be low, and it is difficult to obtain good results when dealing with complicated nonlinear problems.

    “However, when the obtained representative samples are insufficiently representative, the correct rate may be low, and it is difficult to obtain good results when dealing with complicated nonlinear problems.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    ELM can handle complex problems and be more flexible, but requires sufficient data to adequately train its model to achieve better classification results, resulting in lower accuracy when training data is insufficient.

    “This allows the ELM to handle complex problems and be more flexible, but requires sufficient data to adequately train its model to achieve better classification results, resulting in lower accuracy when training data is insufficient.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    We excluded the first 12 bands and last 12 bands because of the noise, resulting in a matrix of 5 rows and 200 columns.

    “We excluded the first 12 bands and last 12 bands because of the noise. The resulting matrix comprised 5 rows and 200 columns.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The overall accuracy was improved by PSO-OL-ELM, indicating that the new algorithm is efficient.

    “Thus, it is evident from Table 3 that the overall accuracy was improved, thereby indicating that the new algorithm is efficient.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The growth rate of AS with high error rate is large, indicating that the recognition ability of this category is improved.

    “Combined with the results of Table 1, it is indicated that the growth rate of AS with high error rate is large, indicating that the recognition ability of this category is improved.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    RDA exhibits strong calculation efficiency and classification capability because of its simplicity and strong theoretical basis.

    “In Table 2, the RDA exhibits strong calculation efficiency and classification capability because of its simplicity and strong theoretical basis.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductnone given

    The size of testing data was small; therefore, the result did not reveal the overall accuracy.

    “In addition, the size of testing data was small; therefore, the result did not reveal the overall accuracy.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    RDA is relatively rigid, but has greater stability when dealing with simple problems and is much more efficient than ELM.

    “RDA is relatively rigid, but it has greater stability when dealing with simple problems and is much more efficient than ELM.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisenone given

    ELM does not need display programming after building the model, while RDA must follow rules given in advance.

    “At first, ELM is a learning algorithm that do not need display programming after build the model and RDA must follow the rules given in advance.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationnone given

    ELM needs sufficient training data to adapt to complex situations, and the generalization performance is stronger.

    “ELM needs sufficient training data to adapt to complex situations, and the generalization performance is stronger.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The correct rate of ELM is not much lower than RDA, and there is still a big improvement after PSO and OL optimization.

    “The correct rate is not much lower than RDA, and there is still a big improvement after PSO and OL optimization.”
    Quote in the paperSubstantial claim
35 of 35

  • The paper relies on this without support.

    Construction waste materials such as plastics, concrete, wood, and bricks can be extremely valuable when recycled or repurposed because they are manufactured via complicated procedures.

    “Construction waste includes materials such as plastics, concrete, wood, and bricks that can degrade the environment; however, these materials can be extremely valuable when recycled or repurposed because they are manufactured via complicated procedures.”

  • The paper relies on this without support.

    These materials are composed of limited nonrenewable resources.

    “They are composed of limited nonrenewable resources.”

  • The paper relies on this without support.

    Hyperspectral data have numerous dimensions that compromise calculation efficiency.

    “In practical scenarios, hyperspectral data have numerous dimensions that compromise the calculation efficiency.”

  • The paper relies on this without support.

    Current hyperspectral classification is executed offline via linear scanning and imaging, which is time-consuming and generates large amounts of data, making online classification difficult.

    “In most situations, this classification is executed offline via linear scanning and imaging. This imaging is time-consuming, and the amount of generated spectral data will increase over time that can be up to several decades. Furthermore, the need for capturing hyperspectral data hundreds of times makes it difficult to achieve online classification.”

  • The paper relies on this without support.

    The accuracy of the classification system is mostly dependent on the classification model.

    “The accuracy of this system is mostly dependent on the classification model.”

  • The paper relies on this without support.

    ELM training yields different results each time because W and B are initialized randomly.

    “For instance, this algorithm first needs to initialize W (weight between input and hidden layers) and B (bias of layer’s neurons), both of which are random. As a result, training will yield different results each time.”

  • The paper relies on this without support.

    The size of the training set affects the accuracy.

    “In addition, the size of the training set affects the accuracy.”

  • The paper relies on this without support.

    The five types of materials are not difficult to distinguish, and do not need too many features.

    “The five types of materials are not difficult to distinguish, and do not need too many features.”
8 of 8

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • ELM citation points to a smoothing paper

    Minor

    The ELM description cites reference 18, which is a paper about smoothing and differentiation rather than extreme learning machines.

  • Smoothing-method citation points to an ELM paper

    Minor

    The text cites reference 17 for the smoothing calculation, but reference 17 is about extreme learning machines, not smoothing.

2 of 2

Cites

Flagged first, then by citations

Research works printed but not linked: 5

  1. R Ma, 2012, The hazard of the construction rubbish on the environment and treatment and utilization, Northern Environment, 4, 33
  2. Q Ni, 2016, Red rosewood wood species identification based on hyperspectral imaging technology, Journal of Zhejiang Agricultural and Forestry University, 33, 489
  3. Guo C. Extreme learning machine and its application in classification. Doctoral dissertation, Zhengzhou University. 2013.
  4. JH Le, 2015, International Symposium on Visual Computing, 682
  5. J Wang, 2013, 45, 100

Cited by

Sorted by citations

Related

Authors
Xiao, W., Yang, J., Fang, H., Zhuang, J., & Ku, Y.
Affiliations
Huaqiao University (CN)
Corresponding author
Yang, Jianhong
ORCID
Venue
PLOS ONE
Publisher
Public Library of Science (PLoS)
Published
2019-01-16
Volume
14
Issue
1
Pages
e0208706
ISSN
1932-6203 (ISSN-L)
DOI
10.1371/journal.pone.0208706
PMC
PMC6334961
OpenAlex
W2910983410
Cited by
27
References
13
Citations by year
2026: 7 · 2025: 4 · 2024: 3 · 2023: 6 · 2022: 5 · 2021: 1 · 2020: 1
FWCI
1.18
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Europe PMC (PubMed Central) (submittedVersion, cc-by)
  • PubMed Central (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
Major Project Foundation of Science and Technology in Fujian Province (2016H6013)
Topics
  • Physical Sciences › Engineering › Media Technology › Remote-Sensing Image Classification
  • Physical Sciences › Chemistry › Analytical Chemistry › Spectroscopy and Chemometric Analyses
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
Keywords
construction waste, waste classification, hyperspectral imaging, industrial cameras, extreme learning machine, particle swarm optimization, online classification, waste recycling
MeSH terms
Algorithms, Industrial Waste, Refuse Disposal, Discriminant Analysis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-16.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-16.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review