Noise-tolerance feasibility for restricted-domain Information Retrieval systems

Information Retrieval systems normally have to work with rather heterogeneous sources, such as Web sites or documents from Optical Character Recognition tools. The correct conversion of these sources into flat text files is not a trivial task since noise may easily be introduced as a result of spelling or typeset errors. Interestingly, this is not a great drawback when the size of the corpus is sufficiently large, since redundancy helps to overcome noise problems. However, noise becomes a serious problem in restricted-domain Information Retrieval specially when the corpus is small and has little or no redundancy. This paper devises an approach which adds noise-tolerance to Information Retrieval systems. A set of experiments carried out in the agricultural domain proves the effectiveness of the approach presented.

Effects of Fe addition on the mechanical and thermo-mechanical properties of SiC/FeSi2/Si composites produced via reactive infiltration

Remaining silicon in SiC-based materials produced via reactive infiltration limits their use in high-temperature applications due to the poor mechanical properties of silicon: low fracture toughness, extreme fragility and creep phenomena above 1000 °C. In this paper SiC–FeSi2 composites are fabricated by reactive infiltration of Si–Fe alloys into porous Cf/C preforms. The resulting materials are SiC/FeSi2 composites, in which remaining silicon is reduced by formation of FeSi2. For the richest Fe alloys (35 wt% Fe) a nominal residual silicon content below 1% has been observed. However this, the relatively poor mechanical properties (bending strength) measured for those resulting materials can be explained by the thermal mismatch of FeSi2 and SiC, which weakens the interface and does even generate new porosity, associated with a debonding phenomenon between the two phases.