SearchUI Options - JohnnyRS/PandaCrazy-Max GitHub Wiki

• There are different options for the SearchCrazy page than the PandaCrazy page. The two options that are the same is the search timer and create search triggers so you can change these values on either page. The general options changes the way the searching will work. Many of those options are the default values to use when creating new search triggers or panda jobs. There is only one alarm that is used for this page for now and can be changed to any sound you want. The advanced options deal with how much memory and database space is being used. Blocked ID's are used to block any HITs or requesters you don't want to do especially for custom search triggers.

---

You can change any of these values. Any time related values will be in milliseconds unless told otherwise.

• [Show Help Tooltips] - This will show the helpful tooltips that can be seen when hovering your mouse cursor for a second over a button, stats or options. This is good to keep on if you're a beginner with this extension. If you feel like you know everything about the extension you can hide the tooltips so they won't get in the way.
• [Search Job Buttons Create Search UI Triggers] - This will change any search jobs created from external buttons to a search trigger instead. This same option can be changed on the PandaCrazy page as well. This is used to slowly move users to use search triggers instead of search jobs.
• [Search Timer] - This value is the time in milliseconds to use for the search timer. It will check the MTURK search page every time this timer is elapsed. Search PRE's are shared with any other script that uses the MTURK search page as well. So if you are using other scripts then be sure to set this value higher or you might get a lot of PRE's. The more PRE's you get the less chance at grabbing a HIT as quickly as you like. Making this value too low could cause problems with MTURK not liking their servers to be bombarded so fast. The minimum set is 750ms.
• [Default Trigger Temporary Duration] - This temporary duration value is used when a trigger finds a HIT and creates a panda job. The default is set to 18 seconds. Every HIT that it finds it will start a panda job collecting for this duration and then stop it but the search trigger will continue looking for more HITs. The reason for this is that a search trigger doesn't have to share time with all of the other panda jobs you have collecting. So a search trigger can find a HIT better but a panda job may still be faster depending on your luck. You can always change this value for any new triggers or change each trigger value to something higher or lower depending on your experience.
• [Default Trigger Temporary Fetches Limit] - Instead of limiting a panda by a time duration this will limit by how many times it has fetched a panda link. Every HIT that it finds it will start a panda job collecting for this amount of fetches. The default is set to 0 because there is a duration already set. You can set this default value here for each new trigger or set it in each trigger's options differently. It's best to either have a duration set or a fetches limit but not both so it won't be confusing.
• [Default Trigger Temporary Ham Duration] - This temporary duration value is used when a trigger finds a HIT and creates a panda job. The default is set at 6 seconds. Every HIT that is found will start a panda job collecting in ham mode for this duration and then go back into normal mode. This can give a better chance at grabbing this HIT. You can change the default value for every new trigger here or set it in the trigger options differently. This value is only used at the start of a panda job that is auto started by a trigger. The value set in the panda job for ham duration will be used after the first ham duration is completed.
• [Default Custom Temporary Duration] - This temporary duration is used for any custom search triggers created. This option is only used when you have auto accept on for a custom trigger. You may keep this the same as the regular temporary duration or have it different. By default it is higher because custom triggers usually is used for a different reason than regular triggers.
• [Default Custom Temporary Fetches Limit] - This temporary fetches option is used for any custom search triggers created. It is only used when you have auto accept on for a custom trigger. You may keep this the same as the regular temporary fetches or have it different. By default it is higher because custom triggers usually is used for a different reason than regular triggers.
• [Default Custom Temporary Ham Duration] - This temporary ham duration is used for any custom search triggers created. It is only used when you have auto accept on for a custom trigger. You may keep this the same as the regular temporary ham duration or have it different. By default it is higher because custom triggers usually is used for a different reason than regular triggers.
• [Page Size for MTURK Search Page] - This is the number of HITS that MTURK will show on the search page the extension uses. The lower the number the faster MTURK can send the search results back so the faster HITs can be found. The higher the number allows more HITs to be shown so the better chance at finding HITs if there a lot of HITs dropped at once. The default is 45 which seems good. For slower connections or computers a lower might be better. This can also lower the internet data usage for users who have a data usage limit. For example let's use a value at 20 which means if 40 HITs get dropped from different requesters then only 20 HITs will be seen by the extension. Usually MTURK isn't that busy so a lower number can work.
• [Minimum Reward For MTURK Search Page] - This price is used on the MTURK search page to show HITs. By default it is set at $0.01 because any HIT at 0.00 usually isn't worth doing but some HITs at $0.00 are for qualifications so this can be a way of finding them. If there are a lot of bad HITs at $0.00 that drops at once then it can have better paying HITs being missed so I recommend it to be higher than $0.00. There is a warning show when changing it to $0.00 so you will understand the problems.
• [Display MTURK Approval Rate For Requesters] - The MTURK approval rate for a requester is shown on the Custom Triggered Hits tab. You may hide this stat with this option if you like. This is a good stat to check before accepting HITs from a requester to see how much they approve others. A lower number should make you rethink about accepting a HIT from them. This is not a stat from Turkerview but something that MTURK keeps track of.
• [Search Page JSON Format] - This option is more of an advanced thing to use to see if there is a difference. It's best to leave it on true so your internet data usage is low and finding HITs is faster. There could be differences from what MTURK sends in HTML format versus JSON format so this can be used to see if it causes more HITs to be found and accepted faster than others using JSON format.

---

Some users may have computers with low memory or disk space so these options will let these users control how much data is saved in memory for faster access and how many history data of HITs will be saved in the database. If you have plenty of memory then you can raise these values so bringing up data will be faster. Data in memory is faster then loading it from the database.

• [Number of Trigger Data To Keep in Memory] - These are the number of trigger data which will be kept in memory for faster access. When a new trigger data gets kept in memory and the amount of trigger data is more than this number then the oldest trigger data that haven't been used will be removed from memory. The minimum value is 20 and the maximum value is 600.
• [Trigger HITs History Days Expiration] - The extension keeps a history of any HITs information that it sees either by a search trigger or just the HITs on MTURK search page. This is done so a group or requester ID can be matched with a requester name or HIT title easily. MTURK doesn't provide some of this data unless a HIT is accepted or on the MTURK search page. Instead of grabbing HIT information from MTURK whenever it doesn't have it I decided it would be better to keep this info for a specified amount of days. This option will let the extension know how many days it should keep HIT data for. The data will be saved in a database on your hard drive so if you have a large space you can keep it for a long time. If you don't have a lot of space maybe a few days will be good enough. Any trigger you have will keep it's data separate from the history database. The minimum value is 2 days and the maximum is 120 days.
• [Custom HITs History Days Expiration] - The custom search triggers will add HITs triggered in the Custom Triggered HITs and it's own history database. It keeps the group ID and requester ID in it's own history and let's the data be kept in the global history database. This value should be less than the history days from the previous option above. I wanted to keep this separate because the custom HITs can find a lot of HITs in a day which could add up quickly. If you have enough disk space on your computer you can keep this high or low if you have low space. The minimum value is 2 days and the maximum is 60 days.

---

For the SearchCrazy page there is only one alarm available to change. When a custom search trigger finds a HIT on the MTURK search page it will sound this alarm. You may Mute this sound or change it to another sound you would rather hear. You can also ask the extension to use a computerized voice to say it found a HIT by clicking the TTS button on. Click the Change button to change the alarm to a sound file on your computer. Select the file with the Browse button and if you hear the sound then you can click the Save Audio button to save it. You can also change the audio back to the default value with the Default Audio button. You may also change the computerized voice which is used to something else on the bottom.

---

Sometimes it is needed to block a HIT you don't like doing or a requester that has no HITs you like doing. This is one way of adding ID's to the blocked ID database. You can add a group ID or requester ID. This is really useful for custom triggers because it will find any HITs on the MTURK search page even if you like it or not. These blocked ID's can also be removed here. This is also useful when you have a RID trigger but you don't like one HIT from that requester. You can block it globally here or block it on the trigger details dialog modal.

• [Adding Blocked ID's] - Once you have the blocked ID's dialog modal opened you will see a text input field at the top with a select input listing any ID's blocked. There are two tabs to separate the group ID's from the requester ID's to make it more organized. You can add an ID by entering it in the text input field at top and then clicking the Add ID button. There is also a way to add ID's to this global Blocked list from the HIT dialog modal buttons when double clicking a HIT in the Custom Triggered HITs tab.
• [Removing Blocked ID's] - Once you have the blocked ID's dialog modal opened you will see a text input field at the top with a select input listing any ID's blocked. There are two tabs to separate the group ID's from the requester ID's to make it more organized. You can remove an ID by entering it in the text input field at top and clicking the Remove ID button. You can also click on multiple ID's in the select input field and then click on the Remove Selected ID button on the bottom to remove those ID's. When you double click on any of the HITs added to the Custom Triggered HITs tab you can also remove an ID from there with buttons.

Search triggers will only see any NEW HITs and won't know about the older HITs. You can force it to check older HITs by first making sure the search page is on and searching. If you enable any trigger it will first check the requester page for ANY HITs they have and will try to collect the first 5 HITs it sees. After checking the requester page it will go back to searching the MTURK search page as usual. If you know a requester has more older HITs you will have to set up a panda job yourself. So for any trigger you want to check for older HITs you can disable it and then enable it again which will trigger the requester page check.

When you click on the details button on a search trigger it will show some options. There are two tabs shown. Trigger options deals with how the trigger will operate and the Panda Job Options deals with the panda jobs that the trigger creates. When a panda job is used it will use those options for each new panda job.

• [Group or Requester ID] - You will see the group or requester ID here for this trigger for reference or copying. This will not show up on custom search triggers because they don't use ID's for searching.
• [Unique Trigger Name] - This is the name that will show on the trigger card for you to know what it is for. This should be a unique name so it's easy to find among all other triggers.
• [Disabled] - This shows if this trigger is disabled or enabled. You can change the value here or just double click on the card to toggle it.
• [Minimum Pay] - This is the minimum pay rate for a HIT to be triggered. For a custom search trigger this must be 0.01 or above. Other triggers can be at 0.00 which means it ignores the pay rate.
• [Maximum Pay] - This is the maximum pay rate for a HIT to be triggered. This value can be 0.00 which means it is ignored. If this value is set then it should be above the minimum value or no HIT will be triggered.
• [Words or Phrases Accepted Only] - This will show the words or phrases that must be in the title or description of a HIT for it to be triggered. I recommend only having one of them so it can search quickly. The more of them added here the more complicated and slower searches will get. Clicking on the Accepted Words or Phrases button will bring up a dialog modal where you can add or remove them to the select input box. Clicking on the Add Word or Phrase will bring up a dialog modal with a text input field. Just enter them here and click Add Word or Phrase button. To remove a word or phrase just select it in the select input box and then click on the Remove Selected button.
• [Excluded Words or Phrases] - This will show the words or phrases that must NOT be in the title or description of a HIT for it to be triggered. I recommend only having one word or phrase so it can search quickly. The more of them added here the more complicated and slower searches will get. Clicking on the Excluded Words or Phrases button will bring up a dialog modal where you can add or remove them to the select input box. Clicking on the Add Word or Phrase button will bring up a dialog modal with a text input field. Just enter them here and click Add Word or Phrase button. To remove a word or phrase just select it in the select input box and then click on the Remove Selected button.
• [Excluded Group ID's] - This will show the group ID's that will be skipped for this trigger only. Other triggers may accept HITs with this group ID so be sure that is what you want. You can use global blocked ID's in the options menu at the top to block ID's for all triggers. Clicking on the Add Group ID button will bring up a dialog modal where you can add or remove ID's to the select input box. Clicking on the Add Group ID button will bring up a dialog modal with a text input field. Just enter your ID there and click Add Group ID button. To remove an ID just select it in the select input box and then click on the Remove Selected button.
• [Temporary Duration] - When a HIT is found from this trigger then it will either create a panda job or start an already created panda job. This option is how long in seconds it should try collecting until it stops. It is used at the start of collecting and not saved in the panda job setting so it is a temporary setting. Usually if a HIT can't be collected for around 18 seconds then somebody else accepted it already so having it on for longer would use up a valuable panda job space. If that HIT gets returned it should show up on the search page again and this trigger might see it and start it again. For some good batches it is good to have this running for a minute or more to make sure you can grab as many HITs as possible. This is because a panda job sees a new HIT dropped before it gets listed on MTURK search page. Each trigger can be very different so this option can be very important to change according to how popular the requester is or the number of HITs dropped per day. The maximum duration is 6 hours or 21600 seconds.
• [Temporary Number Of Fetches] - This option is similar to the duration option above and is only used at the start of collecting a job. This option is not saved to a panda job and is only temporary. This will limit the amount of fetches it will try before it stops trying to collect. It can be the same as using the duration or if there are multiple panda jobs running it could make it last longer because it is counting the number of fetches for the panda job. The maximum fetches is 21600 which means it would last more than 6 hours approximately depending on your panda timer.
• [Temporary Start Ham Duration] - This ham duration is used only at the start of a panda job collecting and not saved to the panda job. Every single panda job created by a search trigger has a ham duration set so if multiple HITs get found it won't try all of them at once. The first HIT it finds will be the one that goes ham and all others will have to wait. If this value is set to 0 then it will use the default value set in the general options.
• [Automatic Collect Hits] - This option will only be shown for custom search triggers. It is used to allow this trigger to automatically start collecting any HITs it finds with a limit. Read more about automation in the Advanced Topics Section.
• [Auto Collect HITs Limit] - This option will only be shown for custom search triggers. This is the number of HITs allowed to be automatically collected when this trigger finds some HITs as long as the Automatic Collect HIts option above is true. Read more about automation in the Advanced Topics Section.

All the following options will be passed to the Panda jobs created.

• [Limit # of Group ID in Queue] - This will limit the number of HITs from a specific Group ID in your queue. For example if a requester drops a dozens HITs at once but only allows a user to submit 2 or 3 from each HIT then this limits the amount of HITs accepted for each HIT.
• [Limit # of Total Hits in Queue] - This will limit the number of HITs from a specific Group ID in your queue if there are a total number of HITs in your queue. This would be good if you want to make sure your queue has space for other HITs that may drop when running a batch with a lot of HITs.
• [Accept Only Once] - Will stop collecting after collecting one HIT. This is usually used for surveys but can be used for anything you want one HIT accepted.
• [Daily Accepted Hit Limit] - This will limit the number of HITs accepted in a day from this job. If you know a HIT only allows users to do 100 a day then this option will limit how many can be accepted per day.
• [Stop Collecting After] - This option is used in the panda job that is created by this trigger. It will stop collecting any more HITs after the specified minutes. This is useful for a HIT that usually drops at a specific time but not after so having it collecting all day would be useless. This is independent from the temporary duration on the trigger option tab and will only be used when you manually press the collect button on the panda job.
• [Stop Collecting After # of Fetches] - This option is used in the panda job that is created by this trigger. It will stop collecting any more HITs after the specified amount of fetches for the panda job. This can be useful if you want to be sure the job has tried a number of times instead of using minutes to stop it. This is independent from the temporary fetches limit on the trigger option tab and will only be used when you manually press the collect button on the panda job.
• [Force Delayed Ham or Collect] - This option is used in the panda job that is created by this trigger. When a panda job collects a HIT it can go into ham mode which means it only tries to collect this one panda job and pauses all other jobs. This is independent from the temporary ham value on the trigger option tab and will only be used when you manually press the collect button on the panda job.
• [Delayed Ham Duration] - This option is used in the panda job that is created by this trigger. It is used for the amount of seconds for a panda job to be in ham mode when a HIT is collected by the panda job. If the previous option is false then this value will be 0 meaning that is not used. This is independent from the temporary ham duration on the trigger option tab and will only be used when you manually press the collect button on the panda job.