See Attached.
Quantitative Methods
- Thread starter Spectre2007
- Start date
What do you mean when you say 'semantic data'?
semantic filters for world news have been used for years--- is this what you mean?
surf
Quote from marketsurfer:
What do you mean when you say 'semantic data'?
semantic filters for world news have been used for years--- is this what you mean?
surf
The financial community does, as you say, quantifies and qualifies news. and has been doing it for years.
I'm sure your Model X has that capability. Mine does too. I rate a Greenspan as 150 DJXX points. I know I will do a bracket and 3 reversals with timing clocks to take advantage of the cascades (See 27FEB07 for recent sematic data cascade data event). Collecting 900 points per contract in the less than 2 hour period is the practical result.
This is an example of the early part of the transition into sematic data.
You can see the "ports" in the form of information data and the technology.
Here, news can be done as a calender. Technology is coefficients and they feed clocks and detectors. Model H has these.
Model B by HDO with 100,000 lines of programming has lines devoted to semantic data like FOMC and the calender, I'm sure.
spectre's Model A may not from whatr he says. He may have a bracket under entry but he probably can't time the cascading not detect it since it is not part of the platforms he mentions.
Above is the obvious and "small talk".
Semantic data is a very well defined field and has many subsets of activity and specialization for making money. The center of activity is the financial services industry and its suppliers web site operators and technology providers. I mentioned these and their languages and the focus of these languages.
Semantic data is created, linked and massaged into very high utility money making information that can be processed by financial trading Models. It is integrated into the Model as a compliment to conventional orthodoxy's usual stuff (See all Models mentioned). This is valuable to the trader. It superior value can be assessed by the usual assessment tools.
For example, when Model H is put into one of the most advanced assessors publicaly available, it produces an effect as well as results. The effect is that the assessment device must have screwed up. The result is past the end of the possibility distribution.
Hybrid Trading Models (HTM's) incorporating semantic data processing "ports" and technology will do the same.
So think of the future as having FA, TA and SA and the integrated versions that will get the higher assessment performance scores.
Consider SA for the 197 sectors of IBD and consider the SA for the 238 sectors of its famous competitor. Both provide FA and TA on the sectors and the subsectors and the individual instruments in the sectors and subsectors.
SA creates vocabularies and definitions. Then these terms are quantified in terms of magnitudes and timing. These data are then deliverable to anyone who sees their value since they are priced as a service.
Take materials as an example. Materials apply broadly to all sectors with emphasis on the non service sectors. everyone knows that materials have markets, etc. The SA part of this is how a given material could work in a potential application because of its chemical nature or biological nature. I have just introduced the "research" aspect of SA as it applies to instruments that are involved in an assortment of sectors. Sector rotation is part of Model H; you can see how I can front run you and Model B which may just be at a level of the "news" semantic example you have in your Model. Certainly I am front running Model A all the time anyway.
SA integrates competitive research as well. It integrates reasearch globally in a context of regulation and environmental and safety and health hoops.
Take dry milling research from a tooling viewpoint. Up to recently only rodent hair and waste were limited in products, like the bread you eat. Now since pets are dying, from the rodent poisons, the poinsens will be intorduced into the regs which will affect the machinery all depending upon the national constraints that are put in place. This is all SA type data and it can be handled with vocabulary, definitions, size and timing, and financial technological configurations.
Sa is just a body of consideration like the present conventional orthodoxy body of consideration. Most models operate with the conventional orthodoxy which HDO describes as the setting in which he operates; the retail version espoused by spectra is also conventional orthodoxy oriented. Model H isn't, however, and it is parasitically related to Models A and B and all the Model X's. It has to be to do what it is designed for. All Models will have to deal with SA as time passes to maintain new creation values that are of a perfomance level that makes enough money.
They do not compete since there is no quantitative methods competition. But all model do have to perform to be practical. They lose practicality as the way the market operates changes. This is hard for people to understand. Often it is just construed that a given methodology comes and goes and that is normal. It is more like an issue of "keeping up with the times".
This is where the words "unbelievable" and "astonishing" get to be used by fringe elements. There has to be a biasic understanding of how the dynamic nature of the financial industry continues to create entropy simply as a consequence of externalities. The two huge impacting externalities as as mentioned. They impact conventional and non conventional approaches. In Model H's case, the workaround on "biggering" of the conventional is already in and it will be changed to be "ported" and have technology and both will go to non existant algorithms in the convnetional orthodoxy.
So now we are in the world of FA, TA and SA.
Your example of news was a good one since it is probably appearnet that not many Models have vocabularies, definitions and quantification of these as yet.
A greenspan is 150 DJXX points....lol..
very interesting, jack. while i understand the appeal and potential of semantic filters, i believe these filters need to be dynamic and constantly tweaked by semioticians due to word flux in this culture. in addition, words that are seen as "good/bullish" can often create the opposite effect. this subject hold tremendous interest for me, however, even with realtime semiotic input, it remains a fixed system/program, and will fail due to the inherent nature of fixed systems and the massive cash hunger of the markets infrastructure.
regards, surf
regards, surf
SA is not fluff stuff.
Think very BIG and think information integration.
Think of a vocabulary that is complex and super.
Then think of being an insider trader as a consequence of two BIG things: You can integrate super quality information that is not now even in the realm of the finncial industry; and you have a technology to use the stuff in manners that are very information and timing effective.
Then think Technology for integrating information and think technology that behaves like what is being used today in the financial industry for the range of present applications but in a manner that it parallels this stuff and deals with quantifying and qualifying the semantically integrated information.
What I am saying is that a parallel universe is coming down the pipe and it will be more powerful than the Models being used now AND it will EMPOWER the present Models so that their measured performance has a NEW VALUATION that more than doubles or synergistically improves things even past that level.
Here is a simple example.
I am tooling away trading and people are coattail trading me as well. Two things are going on. MY Model is getting attention becuase of its performance and I go through the issues of saying it is okay to coattail trade me.
This is well before the WEB and it is before PC's.
The people who are on the cutting edge are doing things to chack out how people trade in retails multiple accounts and they are looking at what is going on.
There are two groups doing this. the brokers who are big and the SEC who monitors the brokers.
How is this happening? Technology is being used.
You can see that. Who has the technology to figure things out? IT IS NOT THE PUBLIC AT THIS POINT.
They figure and I get nailed.
They think I am doing X and I am doing Y.
That is because of the primitive nature of their work and effort.
What I am doing is saying the heck with the mainframes I can work on and I am saying the heck with the 026 card punching.
I am doing a "workaround" on the people who are running the machines of that period.
Today everyone has computers and everyone has Models and they perfom as they do.
Today I have computers and I have another modus that is similar to the offset that I had way back in the example.
when I get cited by the SEC for insider trading and I am just doing something else instead, then I understand that technology and financial information is catching up to where I am operating.
Today, I was trying to explain something NEW. You are talking about things that are present in systems and those things are being owrked around for quite a while also.
SA is something NEW and I will not be able to express what it is nor how it works nor what the "workarounds" are for people who trade in ways to "handle" smart traders (Big or professional).
Give consideration to checking out what is going on out there in the world.
L9ong after I was detected on mainframes of the brokers and the SEC (regulators), the PC came along; the email came along. So did the WEB.
Now SA which means integration of information and the technology for Q and Q it is coming down the pipe as something NEW.
I am not just giving a name to the examples you have posted. I am not talking about facets of HDO and Spec and thier Models; this is NEW and very different and it is going to really wipe a lot of people.
As a retail amateur I had to do the workarounds on the practitioners of the industry to get to high velocity trading. That is just what people like me do.
Other people see that they can operate to service the financial industry to make those who use what they create "have more value created" in their operations. these people are cutting edge types who know how to create advantages for people who are in the game of making money. They are doing a thing called Semantic Data and Semantic Analysis.
They are OUT OF THE BOX of the conventional orthodoxy.
No need to give up an arm. E-signal let's you download tick data for any instrument (you need subscription to the exchange of course) going 10 days back. Within a year or so, you 'll have a pretty good database of very accurate data. Using tick data you can contsruct any chart or run any backtest.