The first year review recommended that WP7 work should focus on the major issues examined in MOLTO, especially in relation to the grammar-ontology interoperability rather than chemical compound splitting. Specific scenarios are needed for the exploitation of MOLTO tools in this case study. It was recommended to include such scenarios in a new version of deliverable D9.1.
In response, two use case scenarios were described: UC-71 and UC-72.
WP7 corresponds to the Patents Case Study. Its objective is to build a multilingual patents retrieval prototype. The prototype consists of three main modules: the multilingual retrieval system, the patents translation and the user interface. This document proposes a methodology to evaluate these modules within the MOLTO framework.
The automatic translations included in the retrieval database have been produced by the machine translation systems developed within the WP5. Hence, the evaluation related to this module is the same as the one described for the WP5 systems.
Nowadays, the IR-facility organizes the TREC Chemical IR Evaluation campaign (http://www.ir-facility.org/trec-chem-2011-cfp) The evaluation campaign has three different tracks. One of them is very related to our objective in this WP. - Technology Survey - Given an information need (from the bio-chemistry domain) expressed in natural language, retrieve all patents and scientific articles which may satisfy this need.
Following the guidelines described in the TREC campaign, the methodology proposed to evaluate the patents retrieval system is as follows.
User interfaces are usually evaluated by means of their Usability. According to the ISO 9241-11, usability must measure the "Extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use.".
Hence, to get a complete picture of the usability, we need to measure the user satisfaction (users reaction to the interface), effectiveness (can people complete their tasks?) and efficiency (how long do people take?).
The three measures of usability are effectiveness, efficiency and satisfaction. They are independent and it must be measured all three to get a rounded measure of usability.
The experiment setting may consist of two scenarios: a closed one (i.e., specifying the information that must be obtained) and an open one (i.e., let the user search any type of information). The users are requested to complete both scenarios, and the order in which they are done must be balanced (i.e., Half of them will do the open scenario first). They must answer the questionnaire twice, just after each scenario.
The potential users might be of two types: MOLTO participants and related people (internal) and external users. The internal users can be used as the control test. External participants can be engaged from tools like the Mechanical Turk Requester [8].