Green Xanax

Buy Xanax online

Showing all 8 results

  • 0 out of 5

    Green Xanax

    Select options This product has multiple variants. The options may be chosen on the product page $276.00$1,156.00
  • 0 out of 5

    Xanax 0.25mg

    Select options This product has multiple variants. The options may be chosen on the product page $270.00$1,150.00
  • 0 out of 5

    Xanax 0.5 mg

    Select options This product has multiple variants. The options may be chosen on the product page $271.00$1,151.00
  • 0 out of 5

    Xanax 1mg

    Select options This product has multiple variants. The options may be chosen on the product page $272.00$1,152.00
  • 0 out of 5

    Xanax 2mg

    Select options This product has multiple variants. The options may be chosen on the product page $273.00$1,153.00
  • 0 out of 5

    Xanax XR 3mg

    Select options This product has multiple variants. The options may be chosen on the product page $274.00$1,154.00
  • 0 out of 5

    Yellow Xanax

    Select options This product has multiple variants. The options may be chosen on the product page $275.00$1,155.00
  • 0 out of 5

    Yellow Xanax Bar

    Select options This product has multiple variants. The options may be chosen on the product page $277.00$1,157.00

๐“‘๐“พ๐”‚ ๐”๐“ช๐“ท๐“ช๐” ๐“ž๐“ท๐“ต๐“ฒ๐“ท๐“ฎ ๐˜ฟ๐™š๐™ก๐™ž๐™ซ๐™š๐™ง๐™š๐™™ ๐™๐™ค ๐™”๐™ค๐™ช๐™ง ๐˜ฟ๐™ค๐™ค๐™ง


Generic name-Alprazolam

Belongs to drug class-Benzodiazepines

Xanax is used for the treatment of anxiety disorders (GAD) and anxiety caused by depression,agoraphobia etc:-

It works by intensifying the activity of certain neurotransmitters in the brain.

It works as a sedative and anti-anxiety drug.

BUT if taken without prescription or overdose may slow/stop your breathing especially if you consume it with various other drugs or alcohol.

Studies And Research About Anxiety?

According to various studies and researches about anxiety, researchers have concluded that anxiety is a complicated issue in this modern world.

ANXIETY is a multifaceted and debilitating condition which include a variety of feelings such as sweating,uneasiness,nervousness,etc:-

Symptoms of anxiety are often seen in adulthood ,nowadays it’s common in teenagers because of increasing day to day competition among peers.

In 2020 a study says 16.6 million prescriptions were written.

Approx 2% to 3%of people in the U.S. suffer from panic disorders,anxiety.

Panic disorders are very common in the U.S.

What does F.D.A (Food and Drug Administration) say about XANAX ?

F.D.A has warned people about the dosage ,which if taken in more the amount prescribed may lead to overdose and death.

Without prescription from a well authorized doctor will increase chances of various health issues if taken either alone or with some other medicines.

F.D.Aย  has provided the most authentic and appropriate information for all Benzodiazepines medicines.

Is Xanaxย  safe and legal?

Answer to your question is YES, it is safe and legal but only prescribed from a well authorized doctor and taken inย  the right amount.

Many individuals take Xanax in large amounts just to be stress free for some time which makes them addictive and they start taking it illegally.Xanax being sedative gives them a feeling of calmness,they feel light,tension free.

Teenagers being financially unstable try to illegally purchase it from unauthorized pharmacies or people who sell them fake Xanax.

Abuse of Xanax :

Abuse of Xanax is common.In 2019, estimated 91.99 million Xanax and other benzodiazepines prescriptions were made in the U.S.

Xanax if consumed with alcohol and other drugs such as anti-helminths ,antifungal,opioids,morphines etc:- can be very fatal even leading to death.ย ย 

ย NEVER OVERDOSE or be taken without prescription.

People who are addicted or are alcoholic try to take it illegally just for the sake of addiction,they consume Xanax in a large amount to feel euphoric.

Are Essential Oils and Therapy useful to overcome anxiety?

ย There are various therapies and yoga asanas which help in overcoming anxiety related problems/disorders.

Ayurvedic traditional practices blend in with various modern /western medicines that are capable of overcoming obstacles caused by anxiety.

Essential oils such as lavender oil,linalool etc:- when used therapeutically can help reduce anxiety.

Avoiding stress by making yourself busy in doing what makes you happy,you can break free from the grip of anxiety and will be capable of overcoming the obstacles caused by anxiety.

Anxiety VS Sleep

Anxiety is nothing but a feedback of stress,a way your brain reacts to our ongoing sufferings.

A normal person i.e without any stress has a normal sleep cycle. They get a peaceful sleep than the person suffering from anxiety or any other mental disorder .

Some people may also suffer from sleep anxiety,which is nothing but a fusion of insomnia and anxiety.This disrupts the sleep cycle which leads to various health issues.

In response to anxiety people may also develop Narcolepsy,ADHD,sleep walking,insomnia etc:-


The number of pharmacies online has increased dramatically in recent times, from US $29.35 billion during 2014 and expected to reach US $128 billion by 2023 across the world. While genuine internet-based pharmacies (LOPs) focus on providing a channel of convenience, and could lower prices for patients,

illegal pharmacy websites (IOPs) allow access for unrestricted availability of prescription medications as well as controlled substances (eg opioids, opiates) and possibly counterfeits, creating a huge danger to the supply chain as well as the health of patients. We know very little about IOPs and the process of identifying and tracking IOPs is difficult due to the sheer quantity of pharmacies online (at minimum 30 to 35, 000) and the fluid character of online pharmacy (online pharmacies can open and close quickly).


This study is designed to rise our knowledge of IOPs by using web data analysis and to propose a unique method of with hyperlinks to forecast and recognize IOPs which is the first step towards tackling IOPs.


We first collected data on web data on engagement and traffic to examine and compare how customers use and interact with IOPs and LOPs. We then developed a simple yet novel method of predicting the legitimacy of pharmacies online (legitimate or illegal) by analyzing the hyperlinks between websites. Based on the framework we created two prediction models, that of the prediction model for reference ratings (RRPM) and K-nearest-neighbor-based reference.


We discovered the direct (typing URL) or search as well as referral are the three most important traffic sources, generating more than 95% of traffic to both IOPs as well as LOPs. It’s alarming to know that direct is the second most important traffic source (34.32 percent) for IOPs. Tested on the data set of 763 pharmacies online and 763 online pharmacies, the two RRPM and R2NN did well, with an accuracy greater than 95% in their forecasts of the status of online pharmacies. R2NN was superior to RRPM in the full performance indicators (accuracy, specificity, kappa, and sensitivity). In the process of implementing both models using Google searches outcome for drugs that are popular (Xanax [alprazolamand OxyContin and opioids) the models generated the error rates of 7.96 percent (R2NN) in comparison to 6.20 percent (RRPM).


Our models for prediction use the information we have (referral hyperlinks) to deal with the many unexplored elements of IOPs. They offer a variety of applications for patients and social media, search engines payments companies, officials or policy makers and pharmaceutical companies to benefit combat IOPs. Since there is a lack of work in this field we are hoping to benefit tackle the present opioid crisis through this lens and encourage further research into the crucial field of drug safety.

Keywords: pharmacy online Web analytics, classification illegal online pharmacies, online traffic analysis

Go to:


The number of online pharmacies (OPs) have seen a tremendous increase over the past few years, growing from US $29.35 billion as of 2014 to US $128 billion by 2023 across the globe, with an annual rate of 17.7 percent 11. The majority of consumers use OPs due to cheaper prices 2., 3as well as convenience and access to a variety of medications, such as those that are recalls or when there is a shortage 4,, 55. But, there isn’t enough knowledge about the prevalence of illegal internet-based pharmacies (IOPs) that are estimated to be between 67% and 75 percent of the drug dealers on the internet 66. While enough work has been done to curb prescriptions in the opioid crisis of recent times However, numerous IOPs add access to prescription-only drugs. IOPs prepare the freedom to access prescription medications, as well as controlled substances, creating enough concern about counterfeit drugs, substandard products and the integrity of supply chains The integrity of supply chains is at risk 7., 88.

Combating IOPs is essential to safeguarding patient safety and the integrity of the supply chain. However, it can be difficult. There is a lack of awareness of the legitimateness of OPs in consumers 9We are still learning about IOPs 66. IOPs might look remarkably like LOPs and unlike other products sold to consumers many consumers do not have knowledge of the difference between potentially inferior medications even when they receive they are. The second issue is that the process of identifying and tracking IOPs is the first step to combat IOPs isn’t easy because of their sheer size and the dynamism of the issue. According to Legitscript the Legitscript 6] that there are 30,000- 35,000 pharmacies online, and around twenty IOPs are created each day. IOPs are created each time a large number of die each day. Although IOPs could be shut down (more complicated than we imagine since the majority of IOPs use servers that are in other countries than in the United States), they appear together other URLs (eg, the 30,000-35,000 IOPs represent just 2000-3500 pharmacies 66).

Some checking systems for OP classification (legitimate or illegal) are available, however with limitations. Certain of them aren’t advised 10among them are such as the Canadian International Pharmacy Association and Pharmacychecker and have been accused of not recognizing the OPs accurately. The two sources suggested from the Food and Drug Administration are the National Association Board of Pharmacies (NABP) and Legitscript. However, both demand that consumers be proactive in looking for the status of pharmacies. According to a study of 500 customers in across the United States, conducted by the Alliance for Safe Online Pharmacies 95% of respondents don’t know about certification programs 9and even how to verify the status of OPs. Additionally, there isn’t an complete database due to the size mentioned above and the changing character of OPs.

This research will use web analytics to gain a better understanding of IOPs and also to forecast, determine and monitor IOPs together the existing data. This can be accomplished in two steps. We first conducted an analysis of traffic based on data collected from web sites, that analyzes the ways in which IOPs and LOPs can be visited and how engaged customers are in the sites. Based on the data from the initial stage, particularly the analysis of referral and referrals data in the third stage, we developed an innovative framework for predicting the legitimacy (legitimate or illegal) of OPs that are based on websites that refer them to. In the framework we designed two easy-to-understand prediction models: the K-nearest-neighbor-based on referrals (RKNN) as well as the method for predicting referral ratings (RRPM) and tried their efficacy together an OP data set comprising 763 Ops. We then applied the 2 methods in Google results. We then compared the two payoff for three well-known substances: Xanax (alprazolam), OxyContin and opioids. These strategies have a variety of uses for shoppers on the internet as well as for others to benefit combat IOPs, as described in the subsection Applications and Conclusions of the Discussion section.

Go to:


Data Sources

We obtained the complete list of IOPs and LOPs from NABP (Legitscript wasn’t available to the extent of our study). NABP offered a list of around 1000 IOPs as well as 50 LOPs. We removed the majority of IOPs which stopped operation when data collection. We then gathered usage data (ie engagement, traffic and details) in the rest of OPs through Similarweb and also obtained information about the structures (ie referrals, backlinks, and data, which will be detailed later) through SEMrush. Since Similarweb doesn’t have information for websites that aren’t in its database, or which have traffic that is not enough to track, this brought to our final size of the sample for each database that are listed in the table 1. The first 5 rows of the table are information about the use of the site, and the last row is structure data. We gathered information from Similarweb via the web by crawling together R. To use SEMrush We efforts to gather the data by hand (no crawling is allowed in SEMrush). If that wasn’t possible we bought the function from SEMrush (it sells various levels of functions via various price-level accounts).

Traffic sources serve the percentage of sources by which users access the OPs such as direct or search, referral displays, social media and email (details in the future). Engagement data provide the amount of the users’ engagement with the site (eg the amount of pages viewed, and the amount of duration of time on the website). Data for countries serve the proportion of the traffic that goes to OPs from various countries. Social media data include the amount of traffic generated by 26 social media sites, like Facebook, YouTube, and Google Plus. Data on search deliver the proportion of traffic that comes from organic or paid search results for OPs. A natural search, also known as natural search, delivers outcome that are provided by the search engine based on their relevancy to the query of the user. Paid results of search payoff are similar to ads that websites pay search engines to advertise their websites for specific keywords. Referral information bring the various websites that refer to pharmacies online and their Internet Protocol addresses, as well as the countries of the country of.

OP Status Prediction Model

One of the challenges in forecasting how the status of an OP lies in the method proposed and the information it relies on should be in a state that is not easily altered by IOPs to influence the future predictions outcome. To address this issue we propose a unique structure-based model that predicts the status of a pharmacy based on the relationships between the websites that refer to the pharmacy. In essence, we believe that if a particular pharmacy is mostly accessed via websites that link/refer to illegal pharmacies, the pharmacie is likely be illegal. Figure 1 shows a simplified illustration of this concept as well as the hyperlinks between the websites of referral.

To implement this concept that was based on the truth list of IOPs and LOPs from NABP We identified all websites that refer to these OPs in the set of data and gathered the structure information, that refers to referrals as well as information about the number of backlinks for each OP in which a backlink is a hyperlink from one website to another (eg an OP here). OP in this case). These information, as shown as in the table 1 as the data on referrals was then used to build the model for prediction. Table 2 gives a summary of these data (the entries show numbers of links linking from the site that refers to a pharmacy).

The pink nodes are IOPs, while the green nodes are the LOPs while the blue and pink nodes represent their referral websites. The diagram shows two fascinating phenomenons: (1) LOPs and IOPs are clearly distinguished by the websites that refer to them (although certain referral websites are able to refer to both IOPs and LOPs) This means that IOPs are typically mentioned by websites that refer to other IOPs and vice versa, and (2) the desirable referral sites tend to form together in groups that refer to their respective pharmacies that are referred to by others while bad referral websites spread across (they use the term “all” to refer to different sorts of pharmacies in between). This 2 phenomenon, particularly the first, support our fundamental idea about together an assessment of the performance websites referred to (ie how often websites that use to refer specifically to LOPs) to establish the status for the OPs.

Based on our model, we believe that the higher Ri is higher, the more likely it’s legitimate. To build our prediction model we established an acceptable threshold that T was the level that was higher than which we believed that the pharmacy would be legitimate. When determining T, we looked at a key aspect, namely the sensitivity of this model. It is the percentage of IOPs that were classified as legitimate. While predicting a wrong pharmacy or in the wrong way can be risky in terms of safety from a consumer’s viewpoint, identifying an illegal pharmacy as legitimate could be more harmful than labeling an real pharmacy as illegal. In light of this it is possible to conclude that:

Step 3: Establish the threshold T of the training set to

so that we consider pharmacy I as illegal in the event that Ri.

Note how this threshold is cautious threshold, which means that a drugstore is likely to be a legitimate one if it is believed to be legitimate. This can affect the precision of the algorithm. In order to maximize the accuracy of the model it is possible to choose an alternative threshold. Another approach we efforts was to test various thresholds with the training set, and then select that one with the best accuracy. Then, we published the accuracy of both thresholds.


Alongside RRPM We then adapted one of the most well-known classification techniques, called K-nearest Neighbor (KNN) which we then incorporated into our framework, based on referral links to create a different prediction model. KNN is an unsupervised model of learning which classifies the sample in the test set according to their proximity to the sample of various classes in the training set 11, 12]., 12[ 11, 12]. The most important aspect of this approach is to define the concept of proximity (similarity). The idea we came up with of the framework we proposed into the definition.

Step 1: Calculate Step 1: Calculate the Euclidean distance of the pharmacie that is x (the that whose condition has to be determined) as well as all online pharmacies that have a status i = 1,2 ,…,n and as

The smaller Di is the more similar pharmacy x appears to pharmacy i with respect to the websites that direct traffic to them.

Step 2: Buy from the pharmacies online in order of decreasing in relation to Di. Note what the standings of best K pharmacies. Based on the KNN tradition the status/class of x is assigned the most frequent status of the K pharmacies. For formal purposes, let the count of legitimate pharmacies in the most prominent K be Ks, and the illicit ones be called Kr. We are aware Ks+Kr=K. Let

x can be expected to be valid the probability of x being legitimate is Rx>0.5. Rx is comparable to the score for reliability of x which is a sign of the accuracy that the predictions are based on, greater Rx means a better prediction.

Like KNN in performance, the RKNN model varies with different amounts of K. Naturally, the too high or small value for K can affect the precision that the model. We tried K=1, 2 ,…9 and reported on the results that the model provided for every K value.

Go to:


Traffic and Use Analysis of the Online Pharmacies

All websites’ sources of traffic include direct or search, referred to as display, social media and email. Particularly, the traffic generated through users directly typing into the URL of the site is classified as direct; while search refers to traffic from search engines like Google, Bing, and Yahoo as well as traffic that comes from sites that have links are considered to be referral, while social is the social media traffic like Facebook and Twitter Display indicates traffic generated by banner ads and Email is the source of traffic from email messages that contain links.

Table 3 provides the mean percent of traffic coming from each source to IOPs and LOPs of our data set on traffic (the standard deviation can be seen in Multimedia Appendix 1). In the table the sources of direct searches, direct, and referral are the three main sources of traffic, generating more than 95% of traffic to both IOPs and LOPs. The high proportion of direct traffic suggests that the OP site is a reputable image and the people who visit the site know what they’re looking for. While LOPs are the most popular opportunity to access via direct traffic (42.48 percent) however, it is troubling to learn that direct traffic is the second-highest source of traffic for IOPs (34.32 percent). This means that customers who have had previous experience using IOPs (eg from searches or referrals) might be able to become regular customers, but without being aware of the risks. It is therefore essential to educate and inform to patients, and stop them to avoid together IOPs on their first attempt.

Research has previously indicated the existence of IOP content on a variety of social media websites 33. Our results show that proportions of traffic coming from social media is less than 5percent for both IOPs as well as LOPs likely due to the fact that a large portion of Ops in our sample don’t receive access to social media. We focused on the LOPs of 24 and 41 IOPs in our sample with significant traffic coming from the 26 social media websites to further analyze and discovered that 92 percent (24/26) of sites that were studied redirect traffic to IOPs while only 50 percent (13/26) of these redirect traffic towards LOPs. While 42 percent (11/26) of the social media websites funnel traffic to both LOPs as well as IOPs, 50percent (13/26) of them redirect traffic only to IOPs. Only 8 percent (2/26) of these websites send traffic only to LOPs. Of the different platforms ( Table 4) We discovered that Facebook is the most effective in directing number of visitors to both IOPs (58 percent) in addition to LOPs (42 percent) and LOPs (42%) – far more than the second-highest (Reddit) which redirects 20% of the traffic toward IOPs while 15 percent to LOPs.

Additionally, our country data ( Table 5) indicate traffic from 52 countries to the 155 pharmacies online which we were in a position to collect data from countries. The traffic from 27 (52 percent) countries is attributed to IOPs. Only 3 (6 percent) countries show traffic exclusively to LOPs. Additionally to this, America is United States is the main user of online pharmacies, accounting for the highest percentage of traffic to both IOPs (97 percent) in addition to IOPs (60 percent) in all countries.

This indicates that the average monthly view (in millions) as well as the number of pages visited as well as the amount of time spent on sites of LOPs are all greater than IOPs. however, it is evident that the bounce ratio (the percent of people who leave the site after only viewing the first page) is less for LOPs than for IOPs. This suggests that when users visit OP sites, they are likely to be more involved with LOPs compared to IOPs. However, there are significant variations between the views per month of websites, which reflects the massive variations between the sites in IOP as well as LOP. For instance among the legitimate ones, and are certain to be the top two, whereas other sites only get just a handful of visitors. Additionally, even though the time spent on the sites of LOPs (5 minutes) was considerably higher than IOPs (3.3 minutes) with P<.001, the most time spent on IOP sites (17.4 minutes) was significantly more than the time spent on LOPs (10.6 minutes). This suggests that if consumers find themselves interested in an IOP, they might be very involved with an IOP and could result in potential transactions. Therefore, it is essential to inform patients prior to they make a decision to enter an IOP, with the prediction method we propose.

OP Status Prediction Models

We present our findings on the accuracy of RRPM as well as RKNN models for prediction. We look at four performance indicators such as accuracy, kappa sensitiveness, and specificity. However, sensitivity refers to the proportion of IOPs that are correctly identified, specificity refers to the percentage of LOPs properly recognized. It is clear this Type I error=1-specificity and Type II error is 1-sensitivity. As we have previously discussed, we decided to use the threshold T in order to achieve an error of type II that is minimal which is the highest amount of sensitivity. Additionally, the proposed model must have a high accuracy and acceptable kappa values which measure the degree of degree of agreement between the class predictions and the observed ones, allowing some degree of the possibility of agreement due to chance 13[ 13 ].

the performance measures the performance metrics RKNN (with the K=1-9) as well as RRPM . It is apparent that all RKNN models have achieved 100 percent sensitivities. However the accuracy, specificity as well as kappa rise but then drop when K rises, and R2NN achieving the best and displaying impressive metrics. RRPM is also able to perform reasonably well and has a sensitivity of 99.2 percent, but with smaller values for kappa and specificity. When you alter your threshold value for RRPM from its current prudent value to the reliability score that maximizes the accuracy of the model within this training dataset, the model’s accuracy, kappa, as well as specificity all rise greatly. However, sensitivity decreased slightly in line with what we would expect.

Implementing RRPM and RKNN on Google Search Results

Our analysis of traffic patterns revealed that searches account for the largest amount of number of IOPs that are visited (39.27 percent). The prediction model we developed can be applied in a variety different ways for the search engine: (1) it can be added to results from search outcome to filter or flag results in search outcome which are likely to be IOPs, and (2) reliable scores for OPs can be used to rank outcome so that the most trustworthy OPs are first. We therefore tested our prediction model on Google results. We also tested it on Google outcome for three popular drugs.

“Xanax” (alprazolam) is form of benzodiazepine. Over 30% of all overdoses that involve opioids also contain benzodiazepines 15[ 15]. There is evidence that suggests that such substances are the most frequent targets of IOPs. We looked at the most searched-for keywords that lead visitors to OPs and discovered that the those keywords that have the drug’s names drew more traffic than keywords that did not have names of drugs. We therefore chose to purchase Xanax on the internet as a keyword and then gathered the most popular 100 results. outcome in the keyword search on September 9th on the 9th of September, 2016. The majority of results of the search outcome included pharmacies that sell Xanax online. In light of the current opioid epidemic, and also the possibility of buying Xanax online, we examined the results of the payoff of the terms “buy” opioids online and buying OxyContin online on the 22nd of April on, 2017. OxyContin has a boxed warning and is a Schedule II controlled substance with potential for abuse like the other Schedule II opioids.

To verify our results We collected by hand the status of OPs from the top 100 results of a search outcome from both the NABP Legitscript and Legitscript. This report provides the status of both sources. The payoff indicate that neither has a comprehensive database, but Legitscript (which is able to check 10 pharmacies per day without cost) has a larger database, which confirms the findings of Mackey who, by hand or through an the automated search for sites related to opioids outcome in sites that are not included in Legitscript. Legitscript database. It’s alarming to know it is not the case that all of these pharmacies listed in the top 100 results outcome are legitimate as per the definitions or NABP nor Legitscript. Then, we used RRPM along with RKNN to determine the legitimacy of these pharmacies. We also we compared our predicted outcome with the status of the OP compatible NABP and Legitscript and NABP

Our model relies on referral data when the referral information for a specific online pharmacy isn’t available and its status is deemed as unresolved according to our models. The study compares the predicted outcome that are derived from RRPM and R2NN and R2NN with the results that are derived from Legitscript as well as the NABP database (NABP numbers are listed within the parentheses) for the pharmaceuticals retrieved by analyzing the best 100 results for search payoff for three keywords (hence 300 overall). For example, of 104 (27) pharmacies have been accurately predicted to be illicit, while two (0) are not correctly predicted as legitimate pharmacies using RRPM when compared to the status outlined in Legitscript (NABP). Additionally seven (0) IOPs alike to the Legitscript (NABP) database can’t be determined by RRPM due to there is no the referral information.


Comparison With Previous Work

In this research we carried out a web engagement and traffic analysis of IOPs and LOPs. We then constructed simple models to predict the condition of OPs that are based on referral links. We then tested the predictions models using information from 763 pharmacies online. While the literature previously available shows evidence of selling drugs via IOPs, there is little research on the web traffic that these sites receive. One exception is the research conducted by Mackey et al [ 5who estimated the traffic to an IOP by together fictitious ads for selling prescription-only drugs that they designed through social media. However, we gathered authentic data regarding the analysis of traffic as well as the predictive models.

In addition, very little research has been conducted to identify and predicting the condition of OPs. The study conducted by Fittler et al [ 10was aimed at identifying those indicators that indicate IOPs by analyzing 136 of them based on the length of time of the website, its geographical area, display of contact information and medical information exchange prescription requirements and verification of legitimacy of the pharmacy. They discovered that the requirement for prescriptions or the availability of contact information doesn’t correspond to the status of a pharmacy that is illegal, as outlined by Legitscript but the ongoing operation of the website shows an extremely strong connection with illegal activities. They did not come up with an algorithm for predicting.

The process of predicting the status of OPs is dependent on categorizing the various websites into distinct categories. There are generally two kinds of methods: material based and structure that are based on structure. There are also hybrid methods. While material-based classification 17uses the web’s material to categorize websites and structure-based classification focuses on the patterns that are present in the structure of the links as well as the topology hyperlinks that are on the website. For instance, Amitay et al [ 18] utilized structural information to categorize eight classes of websites (eg corporate websites as well as search engines and e-stores) with the accuracy of some classes surpassing 85 percent. The prediction models we use are built on structure which makes it simple to discern the things we try to classify in terms of IOP as well as LOP is a lot more nuanced.

Research on the prediction/classification of LOPs and IOPs is very scarce. The only other research is the study of Corona and co. 19which aimed at creating an inventory that contains OPs with textsual material analysis. Be aware that prediction based on material can be more easily altered than prediction based on structure. For instance, if the prediction is based on specific material being displayed on websites that are accessed, IOPs might alter or remove the material to cause confusion in a way that makes it look similar to LOPs. To address this issue this paper suggests an innovative yet simple structure-based concept with relationships between referral websites to help predict the status of OPs.

Also, when searching in the scientific literature to find general predictions and classifying websites offering fake products (not only drugs) Two studies were discovered 20, 2121. Both employed material analysis generally and achieved the accuracy levels of 86.4 percent 20[ 20] and 88 and 88 percent 2121. The method we developed could be applied to other items, rather than only drugs.


In introducing a brand new method, we are confronted with a number of limitations. For one, due to the lack of known ground truth about the state of online pharmacies as well as the information associated with traffic analysis, we only have a very tiny sample (for certain aspects parts of traffic analytics). We believe that a bigger sample size, if it is available, will rise the quality of payoff and enable more precise analysis. While together Google searches results We also have to deal with several websites whose actual status is not known; therefore the ability to evaluate our methodology with the Google results is not possible. payoff provided in our paper is not complete. Additionally, the existing website data source (SEMRush as well as Similarweb) are unable to prepare trustworthy (or even any) information for websites that are small or that lack sufficient information for the overview of traffic. Therefore, the results of this research are applicable to larger, legitimate and illegal internet-based players. The third step is based on referral data from websites. Our current referral database derived from the information we gathered works fine. However, it is evident that the more extensive the database of referral links is, the better. Although old links won’t hinder performance (they are not utilized) it is important to update the links when more ground truth information is available. In the end, we are focused on developing a unique structure-based predictive framework and making easy models in order to benefit solve an important and practical issue. Advanced models, like the hybrid of structure-based and contexture-based models, could be developed in the near future to enhance and boost the performance.

Applications and Conclusions

Recent research has revealed that illegal online pharmacies are in existence and are widely accessible, posing massive risks to integrity of the supply chain and health of the patient. However, due to the magnitude of this issue (>30,000 IOPs) and the dynamism of online channels monitoring and identifying IOPs is the first step in reducing IOPs is a daunting task. In this research we set out to address this gap with an analysis of traffic to rise our knowledge of IOPs and suggested a novel method of predicting the OP status using referral data. We created two specific predictions model (RRPM as well as RKNN) with this idea. Tests of the models against an array of 763 pharmacies online showed that both models performed very well and had accuracy of 95.0 percent (RRPM) as well as 98.6 percent (RKNN). R2NN beat RRPM in more detailed parameters (sensitivity, the kappa as well as specificity). The two models were tested using Google results for search outcome for three drugs, we experienced errors of 6.20 percent for pharmacies whose real status was discovered alike to Legitscript database with an R2NN model, and an error of 7.96 percent for those together RRPM to predict. However, further testing with more data is in the works (the issue is the insufficient real-time data) however, we believe that our analysis of traffic and the method of using web analytics from referral websites to determine the state of OPs is one of the first studies in the field of drugs and offers an efficient and logical method to track OPs.

The model/framework developed has a variety of interesting applications. For instance, they could be used by social media, search engines as well as web-based markets (eg, Amazon), and payment businesses (eg, Visa and Master cards) to block IOPs or consider the status of online pharmacies when rating the search results or deciding the allocation of advertising or making payments, disqualifying vendors, or informing consumers about the possibility of IOPs. They could also be integrated together with social media and search engines to create an alert system for warning consumers to benefit consumers make educated choices. The timelyness of this work could benefit combat the current crisis of opioids. The government agencies, policy makers as well as patient advocacy organizations and drug companies may utilize this method to monitor, identify the impact of IOPs and to educate consumers.

As this is a major aspect that affects the health of patients and the security of the distribution chain believe this study will lead to more accurate and efficient predictions models, or more applications of the prediction models we developed. In a bigger sense we would like to stimulate further research on other aspects to tackle IOPs. Finally, our literature review also reveals that literature on automatic prediction/identification of websites selling counterfeit products (not limited to drugs) is also very scarce, although selling counterfeit products on the web is a prevalent problem. Our model and framework could be applied to different products and we expect to spur research in this area as well.

Go to:



illicit online pharmacies


K is the closest neighbor


legitimate online pharmacies


National Association Board of Pharmacies


K-nearest neighbor based on reference


Reference rating method that predicts ratings

Go to:


Multimedia Appendix 1

Further details about how to engage and the traffic sources information.

Click here to see.(15K, docx)

Go to:


Conflicts of Interest: No reported.

Go to:


1. ePharmacy Market. Transparency Market Research. 2015. [2017-03-03].

2. Quon BS, Firszt R, Eisenberg MJ. Comparison of brand name cost of medications between Canadian Internet pharmacies as well as major US pharmaceutical chain pharmacies. Ann Intern Med. 2005 Sep 20;143(6):397-403. doi: 10.7326/0003-4819-143-6-200509200-00004. [PubMed] [CrossRef] [Google Scholar]

3. Liang BA, Mackey TK. The prevalence and Global Health implications of social media in direct-to consumer drug advertisements. J Med Internet Res. 2011 Aug 31;13(3):e64. doi: 10.2196/jmir.1775. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

4. Liang BA, Mackey TK. Access to online and security of medications in short supply A descriptive study of the characteristics of internet vendors. J Med Internet Res. 2012 Feb 9;14(1):e27. doi: 10.2196/jmir.1999. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

5. Mackey TK, Aung P, Liang BA. Illicit Internet access to drugs that are susceptible to recall and health risks. Int J Clin Pharm. 2015 Dec;37(6):1076-85. doi: 10.1007/s11096-015-0154-8. [PubMed] [CrossRef] [Google Scholar]

6. The Internet Pharmacy Market in 2016: Trends, Challenges and Opportunities. The Center for Safe Internet Pharmacies (CSIP) 2016. [2020-07-28].

7. Jena AB, Goldman DP. A growing internet usage may benefit to explain the increase of abuse of prescription drugs within the United States. Health Aff (Millwood) 2011 Jun;30(6):1192-9. doi: 10.1377/hlthaff.2011.0155. It is Article from PMC for freeยป PubMed[ PubMed CrossRef[ CrossRef Google Scholar]

8. Jena AB, Goldman DP, Foster SE, Califano JA. Illegitimate online pharmacies. Ann Intern Med. 2011 Dec 20;155(12):848-50. doi: 10.7326/0003-4819-155-12-201112200-00008. [PubMed] [CrossRef] [Google Scholar]

9. Consumers’ Lack of Awareness is a Worry for Online Pharmacy Legitimacy. Pharmaceutical Commerce Magazine. 2017. [2020-04-03].

10. Fittler A Bosze G, Botz L. Analyzing aspects of the safety of online medications over the long-term following-up of online pharmacies 136: illegal online pharmacies that are rogue flourish and last for a long time. J Med Internet Res. 2013 Sep 10;15(9):e199. doi: 10.2196/jmir.2606. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

11. Fix E, Hodges JL. Discriminative analysis. Nonparametric discrimination: Consistency properties. Int Stat Rev. 1989 Dec;57(3):238. doi: 10.2307/1403797. [CrossRef] [Google Scholar]

12. Lihua Y, Qi D, Yanjun G. An Improved KNN Text Classification Algorithm Based on Clustering. J Comput. 2009;4(3):-. [Google Scholar]

13. Khun M, Johnson K. Applied Predictive Modeling. New York, USA: Springer Science & Business Media; 2013. [Google Scholar]

14. Cawley GC, Talbot NL. Cross-validation with leave-one-out that is efficient for Kernel fisher classifiers that discriminate between classes. Pattern Recogn. 2003 Nov;36(11):2585-92. doi: 10.1016/s0031-3203(03)00136-5. [CrossRef] [Google Scholar]

15. Benzodiazepines and Opioids. National Institute on Drug Abuse (NIDA) 2018. Mar 15,

16. Mackey T, Kalyanam JJ, Klugman J, Kuzmenko E, Gupta R. Solution to recognize, classify and report illegal internet-based marketing and sales of controlled substances using twitter together machines learning as well as web forensics to fight the access to opioids via digital channels. J Med Internet Res. 2018 Apr 27;20(4):e10029. doi: 10.2196/10029. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

17. Pierre J. on the automatic classification of web pages. Linkoping Electron Artical Comput Inform Sci. 2001:6. [Google Scholar]

18. Amitay E, Carmel D, Darlow A, Lempel R, Soffer A. The Connectivity Sonar: Detecting Site Functionality by Structural Patterns. Proceedings of the Fourteenth ACM Conference on Hypertext and Hypermedia; HYPERTEXT’03; June 11-15, 2003; Maryland, USA. 2003. [CrossRef] [Google Scholar]

19. Corona I, Contini M, Ariu D, Giacinto G, Roli F, Lund M, Marinelli G. PharmaGuard: Automatic Identification of Illegal Search-Indexed Online Pharmacies. 2nd International Conference on Cybernetics; CYBCONF’15; June 24-26, 2015; Gdynia, Poland. 2015. [CrossRef] [Google Scholar]

20. Wadleigh J, Drew J, Moore T. The E-Commerce Market for ‘Lemons’: Identification and Analysis of Websites Selling Counterfeit Goods. Proceedings of the 24th International Conference on World Wide Web; WWW’15; May 18-22, 2015; Florence, Italy. 2015. [CrossRef] [Google Scholar]

21. Carpineto C, Romano G. Learning to Detect and Measure Fake Ecommerce Websites in Search-engine Results. Proceedings of the International Conference on Web Intelligence; WI’17; September 25-29, 2017; Leipzig, Germany. 2017. [CrossRef] [Google Scholar]

Content Copy Source