Dataworks Enterprise Overview

What Dataworks Enterprise Does

Dataworks Enterprise is a product which allows clients to access all Thomson Financial�s rich data content sets through a single interface. It also allows for the integration of the Client�s own databases and third party databases to create a single consistent means of providing data to their users.

Benefits of Dataworks Enterprise Over Separate API Interfaces

Dataworks Enterprise presents a consistent interface to all Thomson data sources which means that the client can access all these sources without writing to separate APIs for each source. This also allows Thomson to make changes to the way individual data sources are accessed without affecting the Client.

Dataworks Enterprise� facility to integrate other sources of information, e.g. the Client�s own data sources, simplifies the process of providing data in a consistent manner.

Dataworks Enterprise allows other Thomson Financial sources to be added as required without additional development; these sources can simply be switched on and the Client can access the data through the same means that existing sources are accessed.

As new data sets are available within Thomson Financial, these are added to the Dataworks Enterprise and consequently are available to Dataworks Enterprise Clients.

Dataworks Enterprise hides any Thomson Financial legacy systems and data access methods from Clients.

Dataworks Enterprise presents technically up to date interfaces and will provide new interfaces as these become available and are considered appropriate.

Dataworks Enterprise prevents multiple accesses to source databases and systems by caching data internally. An additional benefit of this is that responses to data requests are optimised.

Dataworks Enterprise provides additional �platform� type functionality which is not available through individual APIs, e.g. failover, distributed architecture, caching, etc. These are discussed in more detail below.

Input Sources

The Thomson data content sets which are delivered through Dataworks Enterprise include:

GlobalTOPIC A real-time global equities feed which also provides news
GlobalTOPIC Trader Provides trading interfaces to exchanges and RSPs
ILX A real-time US equities and options feed
Worldscope A global historical company financial information database
Datastream The leading financial information database
News (Dow Jones, AFX, ICV, VWD) Available over GlobalTOPIC
Insider Trading Access to company disclosure data
Charts Historical and intra-day, with extensive technical analysis
I/B/E/S earnings and consensus data Broker estimates and recommendations
First Call Estimates Direct and Research Direct.

The third party sources which can additionally be incorporated into Dataworks Enterprise include:

ODBC, DAO, ADO type databases SQL, Sybase, Oracle, etc.
Excel spreadsheets Dataworks Enterprise provides facilities to consume (including real-time market values), calculate and publish the content of Excel spreadsheets.
Word documents

These too can be configured to appear to the user as a single consolidated �database� with other data sources which are being fed into Dataworks Enterprise.

Output Formats

Normally, data is retrieved using the standard API COM objects and converted to its final form. For instance, Dataworks Enterprise contains modules which retrieve data through the API and reformat it as HTML or XML. Alternatively, the data may be employed to feed some other system or business process. Dataworks Enterprise has been used in this manner using XML, Corba and other market data platforms as well as writing direct to customer databases.

Data is generally provided in the most appropriate format, e.g. when price data is provided in XML, text based data (news stories etc.) is provided in HTML.

When using COM or Corba, each item is provided in two formats. If the item is text, both formats are the same, but for numerical items one format is binary (suitable for calculations), the other is for display (it includes appropriate market norms for displaying the data).

Dataworks Enterprise components can be integrated in a Visual Studio development environment.

Major Components Of Dataworks Enterprise

Dataworks Enterprise is a modular system which can be run on a single PC or distributed across a number of PCs located on a LAN or a WAN. It can also be built onto dynamically as the Client�s requirements change or additional back-up / hot standby infrastructure is required.

The main features of Dataworks Enterprise are now explained in detail.

1. Dataworks Enterprise� Cache

The main architectural module of Dataworks Enterprise is the cache. In a Dataworks Enterprise-based system, the cache runs on each machine which is publishing or subscribing to data. Each module of Dataworks Enterprise connects to the Cache which stores any items of data that has been requested. The cache notifies the client when data is available and when updates arrive (if applicable). Caches may be connected between boxes (see below). The cache ensures that data in use resides a close as possible to the user which ensures that further requests for the same information are serviced as early in Dataworks Enterprise module chain as possible.

Additionally Dataworks Enterprise Cache uses a �lazy caching� system where items stay in the cache even after the request for the item is dropped by the client, so that subsequent requests can be serviced quickly without going back and re-requesting the item from the data source. This is most useful in a Web Server environment where users may return to a page soon after moving away from it.

The Cache is optimised to store data in the most efficient way possible both to save cache space and to minimise delivery time.

The cache also publishes the API to the system as COM objects providing a "built-in" SDK.

2. Remote Client/Server

As stated above, caches can be connected using the Remote Client and Server technology which employs TCP/IP virtual circuits to connect the boxes. Data published by sources on the same box as a Remote Server can be consumed by subscribers on another box connected using a Remote Client. Additionally, many such boxes can be "daisy-chained" together, allowing Dataworks Enterprise system to be tailored to specific user requirements.

Dataworks Enterprise also supports an IP multicast based distribution system called Freeway, where larger (more scaleable) systems are being deployed.

3. Data Publishing Options

Dataworks Enterprise has a number of available ways in which data can be published:

  • Publish and expire � the data has a fixed or elapsed date/time when it is removed from all caches. Any subsequent requests for the same data item go back to the source of the data.
  • Publish and update � the data is published and updates are provided by the data source. The updates are provided onwards to all the caches which have stored the original item.
  • Publish and forget � the data is published and remains in the caches on a permanent basis.

    4. Handlers

    At the head of Dataworks Enterprise network, handlers are used to publishing data from Thomson and third-party sources. Typically, a handler uses a legacy API to communicate with the source and then normalises the incoming data to the standard model of Dataworks Enterprise and publishes that data.

    The standard Dataworks Enterprise API supports the development of specialised handlers using the same COM object model as used by clients to subscribe to data. By this means, handler development is kept as simple as possible, reducing costs to implement and deploy systems.

    Plug-ins are specialised handlers designed to support market data platforms. The main distinction between a handler and a plug-in is that a plug-in support both publish and subscribe. Dataworks Enterprise has multiple plug-ins, e.g. Freeway plug-in.

    5.VADS - Value-Added Data Servers

    Dataworks Enterprise also supports VADS. A VAD is both a subscriber to the system and a publisher. Typically, a VAD takes data from one or more sources, transforms it and represents it to the network. For example, Dataworks Enterprise XREF source takes symbology data from GlobalTopic and stores it on a database. It then allows users to make "SQL-style" queries of the data.

    Navigation and Symbol Matching

    As well as market data, fundamental data, research, analytics and news, the Dataworks Enterprise also provides navigational information and symbol matching.

    In some cases, navigational information is available within a single source. For instance, GlobalTopic market data also provide navigational information that links together related instruments. For a stock the following links can exist:

  • related news headlines,
  • the security (which in turn provides links to all other listings of the same security),
  • the company (which in turn lists all the security types listed by the company),
  • any indices of which the stock is a constituent,
  • any options on the stock,
  • any biggest movers chains that the stock is in.

    For other instrument types such as indices, there are other appropriate links.

    Dataworks Enterprise Cross Reference database (XREF) provides an interface to a database of referential information on the instruments which are available to the Client. The database contains all the available codes (as detailed above) and information on what to use to request data from other data sources.

    XREF allows for matching with other data sources. For instance, GlobalTOPIC uses as its main reference symbol the local exchange symbol with an exchange identifier attached, e.g. the symbol for the Paris listing of Michelin is ML.P. XREF also covers the GlobalTopic code. XREF additionally supports:

  • local code which varies dependent on the exchange that the instrument trades on � CUSIP for US instruments, SICOVAM for French, WKN for German, Valoren for Swiss
  • SEDOL numbers where these are available,
  • ISIN numbers.

    Any of these codes can be used to match with other data sources. For example, if the Client�s internal research is referenced by ISIN, this code can be used to match up to GlobalTOPIC.

    Additionally, searching for instruments by name or part of a name can be done through XREF.

    XREF also provides a facility to respond to requests for updates to the instrument database, which allows the Client to identify what instruments have newly come to the market and which have been delisted.

    At a higher level, Dataworks Enterprise infrastructure supports "Resolution". This is an extensible name mapping system, that allows customers to integrate their own symbologies into Dataworks Enterprise and cross-reference between their own databases and XREF. In the future, Resolution will support all the major Thomson name mappings supplied by TKO allowing the user to navigate between any Thomson data source.

    Typical Application for Dataworks Enterprise Platform

    Typically Dataworks Enterprise is currently used to provide data from a number of Thomson sources with the data being displayed on an Internet web site run by a financial institution for its clients or private investors to provide them with the information that they require to make investment decisions, to execute those decisions and to maintain up-to-date information on their investment holdings.

    The typical sources of data are:

    GlobalTOPIC to provide delayed or real-time prices and news which are displayed on the web site; prices are also used to present portfolio valuations.
    GlobalTOPIC Trader to provide straight-through processing for order execution. Thomson has interfaces to a number of back-office systems and additional interfaces can be developed.
    Worldscope to provide company background and fundamental data which is displayed to assist the user in making investment decisions.
    Datastream to display historical price information in charts.
    I/B/E/S to provide consensus broker estimates and recommendations.